House of Lords
Cohabitation Rights Bill [HL]
Friday 5th June 2026
(5 days, 16 hours ago)
Lords ChamberRead Full debate Cohabitation Rights Bill [HL] 2026-27 View all Cohabitation Rights Bill [HL] 2026-27 Debates Read Hansard Text
British Sovereignty Protection (Chagos Islands) Bill [HL]
Friday 5th June 2026
(5 days, 16 hours ago)
Lords ChamberRead Full debate British Sovereignty Protection (Chagos Islands) Bill [HL] 2026-27 View all British Sovereignty Protection (Chagos Islands) Bill [HL] 2026-27 Debates Read Hansard Text
Artificial Intelligence: Impact on Human Relationships and Society
(5 days, 16 hours ago)
Lords ChamberRead Full debate Read Hansard Text Watch Debate Read Debate Ministerial Extracts
The Archbishop of Canterbury
That this House takes note of the impact of Artificial Intelligence on human relationships and society.
The Archbishop of Canterbury
My Lords, I wish to thank the usual channels for allowing me to hold this debate today and the parliamentary staff who have enabled it to happen.
In the Bible, the writer of the Book of Hebrews says of human beings:
“You made them a little lower than the angels; you crowned them with glory and honour and put everything under their feet”.
God created human beings in His own image, with glory and honour—each and every one of us, regardless of who we are or what we do. We carry an inherent dignity and immeasurable value. This is not in spite of our weakness, vulnerabilities or limitations but in many ways because of and through them. God made us to be relational beings, in need of Him and in need of others, not sufficient on our own.
I start here because, fundamentally, our vision of what it is to be human—of our glorious humanity—must inform the rest of our debate about technology and AI. Pope Leo’s encyclical Magnifica Humanitas begins here too. God made us creative beings, and AI and wider technologies are a remarkable product of human creativity. They have led to extraordinary discoveries and breakthroughs at speeds that we could never have imagined. They have connected us across the globe and opened up endless new opportunities for working, creating, learning and travelling. I now carry vast information, processing power and connectivity potential in my pocket with me every day.
But this extraordinary product of human creativity and the power it places in our hands also raise urgent new questions. What are the implications for our human relationships, for our connections with family and friends? How does it impact on our working lives and the existence or quality of our jobs? What are the implications for warfare, for climate change and for our engagement with information and democracy? Just because we could create something or deploy technology in a certain way, does that mean we should? As CS Lewis put it:
“We all want progress. But progress means getting nearer to the place where you want to be. And if you have taken a wrong turning then to go forward does not get you any nearer”.
Wave after wave of technological innovation is taking place as we speak. The question we should be asking is simple: where are we going? What is our vision for how this technology will serve human flourishing? We are in danger of unleashing AI into our lives and societies without the theological, philosophical and spiritual framework with which to make decisions about creating, controlling, using or directing it.
Above all, we need to ensure that AI is being designed, built, regulated and used to serve our glorious humanity and not to diminish it—to be pro-human. As Pope Leo has said,
“humanity—in all its grandeur and woundedness—must never be replaced or surpassed”.
This poses the question: does AI make human life more human? The question matters for those designing, developing and building the technology as they think about what ideologies and belief systems should underpin the models, for there is no such thing as value-neutral technology.
This matters for Governments and policymakers as they determine what should and should not be permitted and regulated, and it matters for those using the technology. Many feel that AI is affecting them hugely without having any say in the matter. Others feel that the things which make us unique as humans are at risk of being eroded, devalued and replaced by AI as people turn to chatbots rather than other human beings for comfort and wisdom in moments of loneliness, loss, anxiety or pain. How should we adopt AI and what is the right place for it in our relationships, our families and our societies?
If humanity is to be placed at the centre of all thinking and decisions about AI, I would like to suggest that there are three fundamental questions to help us to work out what a pro-human framework for AI would look like, and how it informs practice. First, what does it mean to be human? If God crowned humans with glory and honour, how will AI respect—indeed, cultivate—that sort of dignity and value?
There are many ways in which AI is helping to enhance human dignity and to protect and uphold life. If you want to look to the potential of AI to serve human beings, you need look no further than across the sciences. In nursing, we can see the potential of AI. Nursing and medical care are areas where the value of human dignity is visible in some of the most tangible, practical ways. I do not believe that a robot or an AI model will or should ever replace human beings in some of these settings. Sitting at the bedside of a patient to deliver very difficult news or supporting a woman through the delivery of her baby are deeply vulnerable moments where human touch, eye contact and human emotional intelligence are invaluable.
However, there are many ways in which AI is having a hugely positive impact on healthcare. AI is beginning to make childbirth safer, from automated ultrasound to predictive tools for pre-eclampsia to consistent foetal heart rate monitoring. There are, sadly, other uses of AI today which, rather than enhancing human dignity, are providing new ways of degrading it or violating it. A recent report from Durham University presented evidence that chatbots are now facilitating violence against women and girls, allowing roleplay of incest, child sexual abuse and rape with few safeguards, risking the normalisation and the legitimisation of such abuse. These harms are not simply the result of user misuse. AI platform design choices, policies and governance failures are encouraging and enabling them, and existing regulation is wholly inadequate to prevent them.
My second question is: what are we here for, and what gives our lives meaning and purpose? The Christian faith teaches that we are designed for relationship—with God, with others and with the created world around us. We find meaning and purpose in these relationships and in dignified work, where we can partner with God to see His kingdom come on earth. We are human beings crowned with glory and honour, here to glorify God.
Perhaps one of the most profound areas of purpose is work. We are already seeing record numbers of 18 to 24 year-olds in neither education, training nor employment, and this is only set to worsen as agentic AI starts to come online. One area of the greatest gravity that this debate must address is how we as the political class are going to help young people navigate a rapidly changing world.
Yet AI impacts our sense of meaning in other ways that go beyond work. A fundamental quality of a human being, one source of our well-being and sense of purpose, is our ability to create, imagine, think and invent. God placed humans in the Garden of Eden to look after the garden. He asked Adam to name the animals and species there, an inherently creative task.
On one hand, AI can enhance human creativity and imagination. It puts more information, people, networks and tools at our disposal. It can increase efficiency and take away the burden of more repetitive and administrative tasks. But there are reasons to be concerned that AI is in fact having the opposite effect on our human abilities. There is already concerning evidence about the impact of technology, particularly smartphones, on the human ability to think and create. Mary Harrington wrote an article last year entitled “Thinking Is Becoming a Luxury Good”. It highlights evidence of human brain power decreasing, with adult literacy scores levelling off and declining in the past decade in the majority of OECD countries, and some of the sharpest declines are among the poorest.
Child literacy is also declining. Research shows that children who are exposed to more than two hours a day of recreational screen time have worse working memory, processing speeds, attention levels, language skills and executive function than children who are not. A study has shown that heavy users of AI struggle more with critical thinking as they stop thinking for themselves, and their capability atrophies. The irony is that while AI might make us feel like we are more creative, AI’s inherent nature means that, at scale, ideas will in fact become more predictable, unoriginal and homogenised.
The human ability to think and create surpasses the capabilities of AI. We must work to ensure that our human abilities are given space to grow and thrive—to be the thinking, creative beings that God has made us to be. The Pope put this well last November during a live address to young people at the National Catholic Youth Conference in Indianapolis:
“be prudent; be wise; be careful that your use of AI does not limit your true human growth. Use it in such a way that if it disappeared tomorrow, you would still know how to think, how to create, how to act on your own, how to form authentic friendships. Remember, AI can never replace that unique gift that you are to the world”.
My third question is: what is truth? Pilate notoriously asked Jesus this question as he was being sentenced to death. According to the Christian faith, truth is not something we define ourselves or alter to suit our own personal, political or commercial ends. Truth is embodied in the person of Jesus Christ, and expressed in loving God and loving others. Truth is a fundamental foundation on which our personal lives and societies are built. Without common truth, a flourishing common life becomes impossible.
Generative AI cannot tell right from wrong or facts from fiction. Instead of truth, it produces a statistical echo of what has been said before in material it has been trained on. It reinforces biases inherent in the way that it has been coded, as well as social biases present in the material it has been fed. It also simply invents information. One study found that chatbots did this at least 3% of the time, some as much as 27%.
Even more concerning is that AI can be weaponised by malign actors. It is the perfect tool for someone wanting to create fake news. Its ability to disperse disinformation, discredit legitimate information, censor other information and game algorithms has the potential to distort and rewrite reality—to present fiction as fact, and all with a veneer of objective truth. It has the power to manipulate what we see and what we believe at a speed and scale never seen before. What is often presented as a tool in the democratisation of knowledge could, all too easily, become the tool of the autocrat.
The potential for real harm is still to be fully realised. There is a serious risk that this will lead to a fundamental breakdown in trust across society. The real danger is not our rising gullibility, but our rising cynicism. It is not that we will believe anything; it is that we will believe nothing. If we cannot trust information we see online, then perhaps we cannot trust the people we meet. When the possibility for trust in another human being is eroded, relationships cannot be formed, nor can much of what we do in communities or our society be sustained. Without a common understanding of truth, human relationships and the reciprocal ties which underpin our societal structures flounder. Truth and the trust it inspires, between people and within societies, must be cherished and protected.
It is easy to feel overwhelmed when considering the current and potential impacts of AI on life as we know it. What can be done to ensure that AI serves humanity and does not degrade it? Uninventing AI is not an option, nor would we want to be without the many positive effects that it brings, so what are we to do with technology which places great power in the hands of those who own, control and use it? Power is not inherently wrong, but it carries great responsibility and often great risks for human beings, as we have seen repeatedly through history. Power corrupts, and it takes people of great virtue and moral strength to withstand its temptations.
Archbishop William Temple described a central occupation of Christian social thought as being man’s dignity, tragedy and destiny. I have spoken today of humanity’s inherent dignity, but it is our fallenness—the tragedy—which makes technology’s power so seductive and the risk of its abuse so often our story. In the Christian tradition, there is a call that overrides the lust for power; it is the call to service. The distinctive Christian version of service is sacrifice, and Jesus’s life, death and resurrection is the perfect example.
This leads to the third of Temple’s triad: destiny. Jesus’s sacrifice for us on the cross offers profound hope to all humanity in every season and every circumstance. It also offers for us a model of living a sacrificial lifestyle—one where, with Him at work in us, we choose to sacrifice our own personal ends in the service of others and begin to see the kingdom come on earth. Within this, I believe, lies the hope of our relationships, families and societies. At every level of society—from individuals making decisions about how to use AI personally and with our families and children, to business owners choosing where to use AI in their processes and how this affects their workers, to owners of AI companies choosing what technology they invent and the features or limits they place on this—we can choose to make decisions sacrificially in the service of our common humanity.
We must cultivate the character to deal with the opportunities, challenges and temptations that such powerful technology places in our hands personally and corporately, involving people, communities and civil society in conversations about AI, drawing on the wisdom and insights of faith communities. Technology this revolutionary must not simply be unleashed on our societies; it must be developed with us, and for us, at a human pace with human objectives. Above all, with our common and glorious humanity, we must put people ahead of our profit, convenience or technological progress at all costs and ensure that we harness AI to serve humanity, and to be an extraordinary tool in the creation of a more just, abundant and hope-filled world. I beg to move.
Lord Knight of Weymouth (Lab)
My Lords, I am grateful to the most reverend Primate for securing this debate and for the way she has just introduced it. Alongside the Pope’s recent encyclical, moral leadership in this issue is vital.
I should declare my interest in Century-Tech, Goodnotes and the Good Future Foundation, and as a member of your Lordships’ Communications and Digital Committee. I should also confess that I have two main collaborators. Both are called Claude: one is my wonderful wife, and the other is an AI product from Anthropic. Both are great sounding boards. One is brilliant at advising me, looking after me and bring me joy. The other helped me summarise the philosophical positions on relationships that inform my contribution to this debate.
The genius of the likes of Geoffrey Hinton and Demis Hassabis have combined computer science and neuroscience to create machines that out-compete what Aristotle described as the essence of being human—the rational animal—with AI machines that learn, make choices and simulate friendship and intimacy. Yesterday’s Guardian carried an interview with Joanna Stern, who spent a year inviting AI to do nearly everything for her. Her encounter with a companion AI was most striking. What scared her, she said, was
“putting … your emotions in the hands of the machines”.
She came home and hugged her children, saying she never wanted them to have
“a relationship with a chatbot”.
That is already the reality for many. Male Allies UK surveyed 1,000 boys aged 12 to 16 and found that 26% preferred the attention of an AI companion to real human connection. We must act urgently to prevent this consigning a generation to fester in their bedrooms, spiralling into mental health crisis as their addiction to phones gets deeper and darker.
AI has no ethics, none of Heidegger’s mortality or self-awareness. We must listen to Jack Clark, co-founder of Anthropic, who yesterday called for “a brake” on AI. There is a real risk of superintelligent AI systems being developed that could act autonomously from human control, learn their own language to collaborate with each other and present an existential threat to our species. Big tech and AI are already accelerating a breakdown of trust and a rise of division. The attention economy pulls us into echo chambers and away from human-edited media. It is simple to create fake content and distort truth. It is now easy for foreign actors, hackers and shameless populists to spark outrage, then protests and violence, as seen this week. We must make it as easy to be pro-social with these tools as it is to be anti-social. We must do something about addiction by design.
Then, there is education and work. The Milburn review is a wake-up call: there are 1 million NEETs, 7% of 14 to 16 year-olds are unfit for work, and there is a growing mental health crisis among the young. Our obsession with testing in schools is creating a system that alienates too many people, training them to absorb content and produce predicted answers, much as we train AI models, and which AI does better. Young people know that, and they are losing hope as they lose their place in the labour market, whether or not they passed their tests.
AI can help teachers and release them to focus on the human-to-human aspect of their endeavour. But curriculum, pedagogy and assessment must change to nurture empathy, judgment and care for oneself, for others and for the planet. We do not yet know the impact of AI on work, but we know that work is really important for giving people routine, purpose and positivity as social animals. However, we can predict that intelligent robots and software will take over more screen-based work, and more production and service delivery. I suspect that we now need people who can thrive in a relational economy. We will value the personal when we are buying. We already crave authenticity, live performance, the kitchen table at the restaurant for those who can afford it, and the bespoke service or product. That is where the future lies. Future work will not be so much in data centres or in software development; it will be in human contact, unique service and quirky creation. We need an education system that stops standardising humans and nurtures individuals capable of love, empathy and judgment, who can craft, create and communicate. Then, our society will be resilient to the threat of AI and grasp its extraordinary opportunities.
I make a final comment: all this needs the Government and Parliament to have backbone and regulate AI. There may be some who are worried about what that means for a US trade deal. My advice to Ministers is to put the future of our society, our cohesion and our children first.
Lord Waldegrave of North Hill (Con)
My Lords, like the noble Lord, I congratulate the most reverend Primate on this very timely debate and her powerful speech, which will repay reading carefully. I also am delighted to follow the former leader of the Mendip District Council, who at least knows where North Hill is, and I agree with almost every word of his speech as well.
I am going to be accused, no doubt, by some of my colleagues—perhaps even one or two sitting on my right—of crying wolf, because I too am going to call for regulation. But I remind noble Lords that at the end of the fairy tale there is a real wolf. As recursive self-improvement arrives—it may have already arrived—and machines can design machines more intelligent than themselves without our intervention and our control, we are in a completely new world, pace President Milei of Argentina in the Financial Times yesterday. As my late friend Jim Lovelock, he of the Gaia hypothesis, used to say at the end of his long life, “Of course machines are the next stage in evolution. But it’ll be all right: they’ll keep some of us in a zoo for sentimental reasons”. I am not entirely sure I want my grandchildren to live in a zoo.
As the Pope points out, this industrial revolution, unlike the nuclear revolution, is different. It is more like the first one, because the origins of it all come in the private sector, and the Governments are trailing along behind. Among those founders, there are very good people. I happen to believe, like the noble Lord who has just spoken, that in Demis Hassabis we have not only a near genius but a man with a strong moral compass. The work that he and John Jumper did to solve the protein folding problem with AlphaFold is of huge benefit to medicine.
But then there is a potential wolf. Have a look at Evo 2, also a triumph of huge potential, trained on, I believe, 9 trillion nucleotides. It can read the entire blueprint of life. It is in open source. It has guardrails in it, but there are those who say those guardrails can already be got around, for those who want to seek to weaponise bioweapons. We need the humanities to remind us of the possibilities. The most reverend Primate referred to CS Lewis—have a look again at That Hideous Strength of 1945. Have a look at a rather different sort of writer, Gore Vidal, who imagined in Kalki a mad religious purifier wanting to purify the world of sin with a bioweapon. He wrote that in 1978. We need our humanities departments to help us think about the issues here and not leave it to the techies, wonderful though many of them are. Our humanities departments in universities are under very great pressure at the moment. We need philosophers; we need people like Iain McGilchrist reminding us of the left brain/right brain division. We need the Pope, and we need the right reverend Primate, to remind us of humanity.
We need to move fast. A recent study by Barclays Bank predicts that between 70% and 90% of the jobs in installation and repair, manufacturing, construction and warehousing may be taken by humanoid robots—they will be humanoid, because the world is built by us to fit humans—and that will happen over the next 10 to 15 years. So 70% to 90% of those jobs may be gone.
We in this country, thanks partly to the patriotism of Demis Hassabis, Geoffrey Hinton and others, are players in this world. We have a constellation of start-ups and of people thinking about these issues, which makes us among the principal three in the world—far behind America and China, but significant. Thanks to the former Prime Minister, Rishi Sunak, we are in a good position in relation to regulation, because our AISI is widely regarded as among the best of its kind in the world. So yes, we must encourage our start-ups and, yes, we must find wider pools of capital to support them, as the Government are trying to do, to be fair to them. Yes, we must think of regulation.
My final, and short, point is this. We must not be nationalistic about this technology. We made that terrible mistake with nuclear. A predecessor of mine as Minister of Science, much more distinguished than me, who subsequently sat on the Woolsack, said that it was near treason to select the PWR reactor rather than the AGR reactor, because the first was American and the second was all British. Poppycock. We should buy what is best. We should make the public sector an intelligent and careful buyer, but we should not fail to buy the best, from wherever it is produced.
Lord Alderdice (LD)
My Lords, like the previous two speakers, I thank the most reverend Primate for choosing this topic for her day’s debate. As has already been said, it is quite striking that she chose it, as His Holiness the Pope also chose it for his first encyclical, because it is a matter of great concern.
Before I come to the concerns, I want to emphasise that there is a very positive dimension to all this. Technological development has brought us huge benefits in communication, in medicine, and in so many aspects of our lives. But sometimes we forget that, when we talk about artificial intelligence, we are talking about something knowledge based and cognitive, and human intelligence is something different from that. Human intelligence involves the emotional dimension of us, and it is that very fact which enables us to operate in a different kind of way. So rather than simply reflecting on the challenges and dangers associated with technology—of course, it has brought us many benefits, but also many dangers—as we look at things like the environmental crisis or the potential for nuclear holocaust, we should note that these have also been brought to us by technological development, and we need always to manage and control it.
As we travel across London, we may reflect on the fact that driving in a car nowadays is not much quicker than a horse-drawn carriage was 100 years ago. Even so, we have the capacity for speed, so we have to regulate speed. Why? It is not because speed is bad, but because human beings can react only at a certain speed and a certain rate. We can adapt ourselves only at a certain rate. There are limitations to being a human being, as well as advantages and benefits.
Sometimes, the benefits lead us to difficulties. I was a psychiatrist for a long time and dealt with young people with addictions for a period. One reason that medications become addictive is not that they are unsuccessful but precisely that they are extremely successful, at, for example, getting rid of anxiety. Therefore, people quickly get used to reducing their level of anxiety with medication rather than with psychological mechanisms and with relationships with other people. It is the very effectiveness of one dimension of thinking that makes this dangerous for us.
If we look at the human dimension, we need to understand that, for example, speed and a response, albeit sometimes not a correct response, is not something that human beings function well with. When we lose someone, we need a period of time to manage our grief in relation with others. This is not a component of artificial intelligence. In order that we have the possibility of hope, we need to understand the possibility of frustration and the time it takes to accomplish things.
The parents who create some of the greatest problems for young people nowadays are not only those who do not give them appropriate attention but those who give them too much attention. Winnicott devised the idea of “good enough” parenting. One sees many parents who react immediately to children and respond to every wish that they have, without considering other possibilities, and do not understand that this is harmful to their young people because it does not help them to develop the capacity to deal patiently with problems, reflect upon them, deal with grief and find hope emerging. None of these are qualities that artificial intelligence either has or has the capacity to develop.
I therefore make an appeal that, as we talk about artificial intelligence, we should understand it to be something fundamentally different from how we function as human beings. The more deeply we can understand how we function as individuals and as communities of people, the better position we will be in to address the challenges that arise from artificial intelligence.
It is not just that these are problems for the future. It is absolutely clear that we already face many difficulties that emerge from the fact that young people now look around and see a world where there is environmental threat, where there is the possibility of nuclear war and where they do not understand how they are going to deal with artificial intelligence. We ought to reflect on the epidemic of adolescent mental health problems, ADHD and autism. Are these things not related to anything? Are they only better diagnoses, or are there things happening that are making life difficult for the next generation?
Baroness Kidron (CB)
My Lords, I too thank the most reverend Primate for her wonderful introduction, and indeed acknowledge the intervention of His Holiness Pope Leo, who I had the privilege of meeting at the Vatican on this subject some months ago.
The relationship between humanity and AI is often described through the language of alignment, as if it were a technical requirement. An AI system is considered aligned if it advances the intended objectives and misaligned if it pursues unintended consequences. However, before we ask whether a machine is pursuing the intended goals, we must first ask who gets to define those goals. That is not a technical question, but a political one. AI is brilliant—it is able to identify cancers, anticipate environmental disasters, accelerate scientific research and improve productivity. It will transform human life. But technology is never neutral. Every technology reflects a decision about who benefits, who bears the risks and who gets to decide. Every other industry, from pharmaceuticals to aviation, and from nuclear power to the sewer system, is regulated to ensure that private gain is balanced against the public good. Yet the tech sector has spent 30 years arguing that it should be the exception and it continues to make that argument about AI.
We have already seen the cost of tech exceptionalism on the bodies and minds of our children, and in our public discourse, deserted high streets and the weakening trust in our democratic institutions. Those outcomes were not technical errors; they were predictable consequences of a business model designed to maximise engagement, capture markets and minds, and create indispensable, if unreliable, intermediaries to all aspects of human life. The systems were aligned perfectly. They did exactly what they were designed to do, but they were aligned with commercial incentives rather than the public interest, or, as Pope Leo would put it, the common good. That matters, because many of the same companies, investors and assumptions are now shaping the next generation of intelligence systems.
While much of the public debate focuses on future dangers of superintelligence, joblessness and the world of unprovable truth, it risks distracting us from the more immediate. AI is already shaping our economy and society. It influences what we see, what we believe and, increasingly, what opportunities are available to us. It is already transferring huge sums of money from the UK to Silicon Valley, while concentrating control over increasingly essential infrastructure in the hands of a very small number of corporations. We are living through an extraordinary moment in which perhaps the greatest technological opportunity in human history is also becoming one of the greatest transfers of power and wealth in history—a heist in plain sight.
Our religious leaders are asking us to judge artificial intelligence not by what it can do but by whether it serves human dignity and the common good. That sounds simple, but it is truly radical. It means asking political questions about how we deploy technology. Does it distribute power? Does it strengthen families? Does it create dignified work? Does it increase human agency? And, ultimately, who are the beneficiaries?
Beyond that, it demands political leadership. In this House, we have tried to provide that leadership. We have argued that citizens should benefit from the value of their creative labour, that the data in our public institutions should be treated as a sovereign asset, that computer-generated evidence should be reliable and that AI systems should be tested to ensure they cannot create child sexual abuse. We have argued that chatbots should be subject to the law like every other actor in society. That is what alignment looks like in practice: not a technical specification but a democratic one. I ask the Minister to explain why the Government have repeatedly rejected measures designed to secure democratic oversight, accountability and sovereignty, and what plans they have to ensure that we live under the rules of UK democracy and not those of Silicon Valley terms of service.
Baroness Bottomley of Nettlestone (Con)
My Lords, it is a great honour to speak in this debate. It will sound increasingly like an echo chamber in the themes that are raised but, to follow the noble Baroness, Lady Kidron, with all that she has done in this sector, is remarkable.
I start by paying tribute to the most reverend Primate the Archbishop of Canterbury. She gave a magnificent speech. I speak on behalf of many, I believe, in saying how delighted we are that she is the Archbishop of Canterbury. She learned humanity at the bedside. Those of us who have been involved in health know that it is the nurse who really is closest to the patient at times of suffering, difficulty and frailty. That shone through in her remarks today. We are privileged indeed that she is our archbishop. We all pay tribute to her for the topic of the day, following on from the Pope’s encyclical.
The most reverend Primate cannot have all the glory, however, because the right reverend Prelate the Bishop of Oxford deserves a great deal of recognition for his steadfast tenacity on the importance of ethics in relation to AI and human-centred AI. Indeed, former Archbishop Welby, who made such a big and profound difference for so many of us, supported the Rome call for AI ethics in 2024, so there is a continuity about this.
As the most reverend Primate said, artificial intelligence already offers enormous benefits to humanity across science, education and industry, with the potential to increase productivity, solve complex problems and improve quality of life. We welcome those advances, nowhere more than in healthcare. As the most reverend Primate said, AI is helping clinicians detect diseases, personalise treatments and improve patient outcomes. It is transforming medical research; the work of DeepMind, through AlphaFold, and of Isomorphic Labs, has demonstrated AI’s remarkable potential to accelerate drug delivery by predicting the structure of proteins and helping researchers identify promising new treatments at unprecedented speed. What once took years of painstaking laboratory work can increasingly be accomplished in a fraction of the time, bringing hope to millions. It reminds us that AI is not just a technological advance: it has the potential to become the greatest tool for human flourishing ever created. If harnessed wisely, it can help us live longer, healthier and more productive lives.
Of course, innovation and governance must go hand in hand. This week, noble Lords have been rediscussing the Human Fertilisation and Embryology Act, which was my first Bill when I became a Minister. This was a new technology, creating a life outside the body, and it needed different regulation: subtle, complex and long-lasting. Now, it needs updating. It was difficult to organise that regulation and this is a whole new challenge—but it has to happen. We have to strike the right balance between encouraging the development of technologies that benefit humanity while ensuring that they do not undermine human dignity, security, freedom or well-being.
We need not only to address the problems of AI that are here and now, but to understand and forestall the problems of AI that are to come, which could dwarf those we face today. There are the fast-burn problems of advanced AI in the hands of bad actors or the loss of control of an advanced AI; there are also the slow-burn problems we are just starting to confront, including the erosion of jobs leading to many of people being deprived of a job, career and sense of purpose in their lives, and the threat of gradual disempowerment and enfeeblement of human beings, as has been said. No one knows how to avoid the fast-burn problems, but supposing those potentially catastrophic threats are solved, it is the slow-burn problems we will have to contend with.
Rather than addressing the problems of today—troubling though they are—we must show our concern for the rapidly advancing future. It is a future we have not asked for that is being imposed on us by a few tech barons, backed by massive amounts of money. As the right reverend Prelate the Bishop of Oxford rightly asked in his response to the King’s Speech, where are the laws binding development of superintelligent AI? Those who want to have several sleepless nights need only read Yudkowsky and Soares’ recent book. Do we want AI that is more capable than a human in all significant respects or and AI that none of us knows how to control?
Enlightened governance, transparency and safety measures are not obstacles to innovation; they are essential means. We must address the potentials of the future now. Humanity has a challenge on its hands; it is one we can meet, but only if we combine wisdom with innovation, ethics with progress and a steadfast commitment to ensuring that technology remains the servant of humanity, not its master.
Lord Nagaraju (Lab)
My Lords, I am grateful to the most reverend Primate for bringing this timely debate. I declare my interests as a technology consultant, an adviser on AI policy and a director of companies and organisations involved in AI and AI governance.
I support this Motion because Al is not merely another tool; it is increasingly a mediator of work, knowledge, attention and relationships. Unlike earlier technologies, AI moves at digital speed and can spread at very low marginal cost through existing digital infrastructure, entering our phones, homes, classrooms, hospitals and friendships, automating not only labour but judgment, creativity, companionship and trust.
A small number of firms and countries now command many of the models and much of the compute and data on which others increasingly rely on. This makes AI not only an economic race but a governance challenge. Like all nations, we cannot afford to fall behind. AI will be central to productivity, public service reform and national competitiveness. The Office for Budget Responsibility estimates that around 40% of UK occupations are exposed to AI, with most likely to be complemented rather than replaced. We must therefore be ambitious, innovative and open to enterprise.
Innovation and responsibility are not opposites: trustworthy AI should become one of Britain’s competitive strengths. The dilemma is clear: move too slowly, and we risk economic decline and dependence; move too quickly, and we might embed systems that erode dignity, privacy, fairness and trust. Governance must include technologists and investors, but must not be left to them alone. Ethicists, teachers, parents, clinicians, workers, entrepreneurs, civil society, young people and parliamentarians must all have a voice.
AI may deepen existing divides. Countries and companies with compute, data, talent and capital will shape the terms of use. Communities with access to AI-enhanced education, healthcare and employment may pull ahead, while those exposed mainly to surveillance, misinformation or job disruption may fall further behind. Successful companies create jobs, wealth, tools and opportunity. But democracy must be capable of scrutinising concentrations of economic, informational and infrastructural power moving at this speed.
More than half the people in the UK aged 16 and over use generative AI. Around one in eight users reports using AI as someone to talk to or as a friend—and the figure is even higher in the United States. Evidence on long-term effects is still limited, but the risk is clear. Digital platforms already shape identity and social habits, especially among young people, but generative AI goes further. It talks back, remembers, flatters and can simulate intimacy. Young people may increasingly turn to it for advice, reassurance and help with social situations: roles once shaped by parents, peers and teachers. AI companions do not require reciprocity. Real relationships involve patience, disagreement and obligation. We should not discover too late that human development has been reshaped by systems that simulate care without sharing responsibility.
AI’s benefits are real. In the NHS, AI can reduce administrative burdens, triage referrals, predict missed appointments and support cancer screening. In education, it can help teachers plan lessons and create resources—though it must not weaken teacher-pupil relationships or critical thinking. However, we must also plan for the environmental cost. AI data centres need significant electricity and water. That is not an argument against AI infrastructure, but for building it with resilient energy and water planning from the start.
Some of these questions cannot be answered within our borders alone. In July, the United Nations will host the first Global Dialogue on AI Governance in Geneva, which I hope to attend. The UK should help shape global rules that are innovative, entrepreneurial and at the same time democratic, humane and inclusive. The questions raised by artificial intelligence are not only technical and economic; they are about what kind of society we wish to live in, how work is valued, how children learn and form relationships, and how innovation and enterprise can serve the common good.
I welcome the Government’s efforts to place the UK at the forefront of safe and responsible AI. As that work develops, I hope Ministers will continue to consider AI’s wider social impacts, including on children, loneliness, education, work, public services, competition and democratic trust, and ensure that Parliament remains engaged in that conversation. The Motion before us invites us to take these questions seriously and to bring together expertise from across society. I support it wholeheartedly, and hope this debate contributes to the Government’s wider work, and to the national and international conversation about what AI is, whom it serves and how it can be directed towards the common good.
Lord Ahmad of Wimbledon (Con)
My Lords, I too join the chorus in thanks to the most reverend Primate for her excellent and thoughtful introduction to this debate—and, of course, for the impeccable timing. As we have already heard from the noble Lord, Lord Knight, Anthropic’s co-founder has called for a “brake” on AI. At the same time, we are hearing from Cambridge that artificial intelligence has developed a vaccine. I argue—I hope the most reverend Primate agrees with me—that divine intervention is at work. Putting God at the centre of our thinking, as a person of faith, I also accept that technology shapes our life in the modern age. Yet behind these changes, we should never forget, as we have been reminded by a number of speakers, that it is the human mind that gave birth to that new idea and that technological advancement.
I grew up in a very close-knit family. Discussions around the dinner table were not optional and not merely encouraged; they were vital and part of our fabric. My father likened the advent of the mobile phone age as the start of the decline in the art of human conversation. Calculations, he said, were not for calculators but for the mind—he was an accountant. Yet to my children’s generation, AI is and will become a greater part of their lives. I recall my youngest son—who will be embarrassed by my reference to him—expressing total disbelief that my childhood had no mobile phones or internet. Indeed, he refers to that golden age and my father’s words as “the ancient times when Daddy was a child”.
Behind this humour lies a serious challenge. AI is no longer a distant concept; it is not science fiction. That reality presents extraordinary opportunities and, in terms of defining human relationships, an ever-evolving challenge. In relation to faith, I have seen AI search scripture and provide visual and interactive learning at the touch of a button. AI is changing learning, as we have already heard. Remote villages can access knowledge once the bastion of the greatest universities. In medicine, doctors can diagnose and treat illnesses more quickly and even remotely. Scientific discoveries may now take months rather than years. The potential for human advancement through AI is incredible. To quote Captain Kirk from “Star Trek”, “the final frontier” is being realised not just in space but right here on earth.
Technology has always been most powerful not when it changes operations but when it changes human behaviour. The printing press and the telephone transformed communication. The internet transformed connectivity. AI transforms relationships. That is why we must ask not only what AI can do but what it should do. Human relationships are built on qualities that cannot easily be coded: trust, empathy, compassion, loyalty, forgiveness and love. As the most reverend Primate reminded us, these are not algorithms; they are deeply human qualities fostered by the beating heart, the human mind and lived experiences. An AI system may tell us what someone is feeling, but understanding sorrow is not the same as experiencing grief. Simulating compassion is not the same as feeling it. A machine may generate a sympathetic response, but only a human heart truly cares. Yet we must also acknowledge that AI may help strengthen relationships, as we have heard, for families and for people with disabilities. Used wisely, AI can bring people together; used poorly, it can drive them apart.
We are already seeing a world where individuals can spend hours connected digitally while feeling increasingly disconnected emotionally. The noble Lord, Lord Alderdice, made that point. If a machine becomes easier to talk to than another person, we may lose some of the patience essential to human relationships. After all, unlike AI, our spouses, children, friends and colleagues do not arrive—I am sure the noble Lord, Lord Knight, agrees with me—with “update settings” or “refresh” buttons. Some may seek it. I know my wife, in my absence as a Foreign Minister, probably sought it—unsuccessfully, I would argue.
Perhaps that is precisely the point. As we have heard, human relationships derive their value from their imperfections: the awkward conversation, the misunderstandings, the disagreements and the look. These are not flaws in humanity; they are part of what makes us exactly that—human. We must ensure that innovation is guided by ethics, that regulation keeps pace with capability and that human dignity remains at the centre of technological progress. Therefore, can the Minister say what the Government’s plans are on ensuring that these principles are upheld? What conversations about that are we having with partners?
In conclusion, we must teach future generations that, while artificial intelligence may be able to answer some questions, it is human intelligence, wisdom and compassion that give meaning to our lives. At least, that is the argument that my wife and I present to our three human wonders at home—their response is, “Hmm, we’re not convinced”. The future should be a partnership in which technology and AI enhance the best in humanity, as the most reverend Primate reminded us. If we get that balance right, artificial intelligence will not diminish human relationships; it will give us, perhaps, even more reason to value them.
Baroness Fall (Con)
My Lords, I join the chorus in thanking the most reverend Primate for an excellent and timely debate and a brilliant speech. It is also always a pleasure to follow my noble friend Lord Ahmad.
We are already in the midst of a technical revolution. AI is only the latest iteration to radically change the way we live, work, think and even love. We have learned the hard way, from our battles with social media, that unless we seize the moment and try to mould technology to our value system, we will be left navigating the Wild West. AI is much larger and more consequential. Debate seems to swing between two extremes: the optimists who believe that AI will solve all our problems, and the pessimists who believe that robots have already taken charge. As usual, the truth lies somewhere in between.
This is no futuristic debate—AI is already here—which is why this is such a commendably timely intervention. We can already feel AI in the workplace. We hear business leaders speaking about the opportunities for improving productivity, although they do not really know exactly how, or whether AI will give way to new and unknown jobs. The transition is likely to be a difficult one. However uncomfortable it may be, attempting to put the genie back in the bottle is not a strategy. We in Britian cannot afford to stand back and miss the AI boat, for our own future prosperity and security.
As we grapple unsuccessfully with low growth and low productivity, there are so many areas where AI could be a rocket-booster. We are a country with world-class universities, a strong financial sector, leading scientific research and a thriving creative industry—all areas where AI can drive enormous growth, if we get the policy and investment right. But—here is the crux—the debate over last weekend between former Prime Minister Blair and Burnham and Streeting seems to be a classic misunderstanding. There should be no trade-off between growth and inequality. Growth is vital to the health and welfare of our nation, as is ensuring that those benefits are felt by all communities. These aims must go hand in hand.
It would be extremely dangerous to turn our backs on AI because of concerns around driving further inequality, just as it would be foolish to ignore the possibility that it could. We should not be blind to the danger that the latest tech revolution continues the trend we have already seen in globalisation in delivering for some but not others. When people begin to feel that the economy is no longer working for them, and that they have no stake in its success, the politics of that frustration follows quickly behind.
No group is more exposed to AI than the young, especially those about to leave school wondering what best to study next and whether there will be a job for them when they leave higher education. Alan Milburn spoke devastatingly of the silence of the unheard voice of young people applying for jobs. His recent report has shaken the nation. One million and growing young people who are NEET is not just a national tragedy but an economic one in slow motion. We must ensure that the Government do not make policies that make this worse. It is a moral and economic imperative that these issues are tackled, especially at this time of AI.
The dignity of work is a very human condition. We heard Pope Leo and the most reverend Primate the Archbishop of Canterbury speak so eloquently about what it is to be human. I was struck too by the call to arms. Setting boundaries for AI is a matter for all humanity—no one is without responsibility to shape the future of AI, including and especially those who dedicate themselves to the AI race. In the pride and ambition to win, we must never lose sight of what we are trying to achieve. As the co-founder of Anthropic said this week, morality is not a side business.
We must focus on what we want from technology, before it imposes what it wants on us. There are no easy answers. We are grappling with a revolution in real time. We are trying to make it bow to our values. While the opportunities are clear, it is the responsibility of each and every one of us to push back against the excesses of AI and to grapple with the moral issues it raises, for AI will shape the future, and we must shape it in turn.
Lord Tarassenko (CB)
My Lords, as an AI researcher for nearly four decades and honorary lay canon of Christ Church Cathedral in Oxford since 2012, I congratulate the most reverend Primate on securing this hugely important debate. I declare an interest as the founder and director of LIO, an AI for healthcare company. In the context of this debate, I stress that it is spelled LIO, not LEO.
In November 2022, OpenAI fired the starting gun of the AI arms race by releasing ChatGPT—a large language model with a chatbot. When it released ChatGPT with very basic guardrails only, OpenAI knew that it had major issues. The early release set the pattern in the AI arms race: models released as fast as possible without a comprehensive evaluation of their capabilities and alignment with human values.
This can have disastrous consequences—for example, in the use of AI chatbots by teenagers. Teenagers may start using AI chatbots to help with their homework but soon spend more time with them until they develop an unhealthy dependence on them. This dependence is fed by the sycophantic traits of LLMs, and in extreme cases this has led to tragic outcomes for young people. This has not stopped a third of UK teenagers using a chatbot for an emotional relationship. Some 23% of them actually believe that AI can feel emotions and is conscious.
But it is not just teenagers who are deceived. Richard Dawkins, an evolutionary biologist and prominent atheist, wrote in an essay last month that Claude, the Anthropic LLM, showed
“a level of understanding so subtle, so sensitive, so intelligent that I was moved to expostulate, ‘You may not know you are conscious, but you jolly well are’”.
A dividing line is emerging between committed atheists, such as Geoff Hinton and Richard Dawkins, and those of us from Jewish, Christian and some Islamic traditions who believe that there are fundamental aspects of human nature that cannot be reproduced in silicon.
Despite sophisticated reasoning abilities, AI can produce only simulations of human-like behaviour, acquired through pattern recognition at an astronomical scale, not actual experience or understanding. An AI chatbot cannot experience pain or joy, even if some LLMs, trained on datasets that include descriptions of human suffering and sci-fi stories of sentient machines, can deceive people into believing otherwise. But as Mustafa Suleyman, the CEO of Microsoft AI, wrote earlier this year:
“When a user interacts with an agent, the system should consistently work to puncture the illusion that it is any kind of sentient being … The simulation is getting better every day”.
We have to educate young people about this illusion, so that they understand they are interacting with a machine with human-like capabilities, not another human being. The DfE has a responsibility to do so when it introduces AI tutors into the classroom at the end of 2027. We also have a moral duty to put pressure on AI labs—or regulate, if needed—so that AI chatbots are only released with effective guardrails that prevent human beings forming attachments with them.
Is there a glimmer of hope that we can begin to influence the direction of travel of the AI arms race? One of the co-founders of Anthropic was present at the launch of the Pope’s recent encyclical. He said that every frontier AI lab experiences the
“pressure to stay commercially viable … at the research frontier”,
which
“can sometimes conflict with doing the right thing … We need moral voices that the incentives cannot bend”.
Earlier this year, in a standoff with the Pentagon, Anthropic refused to remove safety guardrails on its Claude models for military applications. Doing the right thing goes beyond military applications, however. The biggest failure of AI labs has been to design AI that is smarter than a human as an end in itself, without proper regard for its potential harms to society, especially young people. As moral voices we can point this out, but we need the AI labs themselves to acknowledge this failure. Then and only then will typical mission statements from these labs about wanting AI to be safe and beneficial for society as a whole become more than just an aspiration.
The Lord Bishop of Oxford
My Lords, it is a real pleasure to take part in the debate and particularly to follow the noble Lord, Lord Tarassenko, with his wisdom and great knowledge of the field.
Humanity stands at a real crossroads in the present moment. Artificial intelligence brings many potential benefits, as the most reverend Primate the Archbishop of Canterbury said, but also great jeopardy for the world. The Prophet Jeremiah invites the people of his generation facing crisis to:
“Stand at the crossroads, and look, and ask for the ancient paths, where the good way lies; and walk in it, and find rest for your souls”.
The primary risk of AI in the present moment is that the technology is being developed primarily for profit, financial gain and shareholder benefit—as the noble Baroness, Lady Kidron, said—rather than for the benefit of humanity and our common home. Those commercial gains are mind-boggling for the few. But as profits rise, the very concept of what it means to be human is being reduced. Human beings shrink in dignity and glory to become simply consumers of goods or services—here to give our attention, money, brand loyalty and image—or else women and men, made in the image of God for life in all its fullness, are reduced to units of production, serving the algorithm and filling in the ever-diminishing gaps left by robots and AI.
Warehouses, call centres, transport, entry-level jobs in the professions and coding are all seeing vast reductions in the workforce in the aim of efficiency. We need a different kind of prophet now. What kind of future are we creating? Is it an empty utopia for some, based on entertainment that will not satisfy and what feels like slavery for others?
Nowhere is this dilemma seen more clearly than in the area of work. Elon Musk famously predicted in 2023, in dialogue with our then Prime Minister, that AI will put an end to jobs, as if that is a good thing. But work is not merely an instrument. As Pope Leo has argued, work
“expresses and enhances the dignity of our lives. It is a requirement of the human condition, a normal path toward maturity, development and personal fulfilment”.
Last year, in the diocese of Oxford, we invited 100 young people, mainly sixth-formers, into our clergy conference to listen to their concerns about the world in small groups. AI and tech were at the very top of their list for a variety of reasons, but the most common was what AI would do to the jobs and careers of these outstanding young people. Those fears have been underscored, as noble Lords have said, by Alan Milburn’s report on the rising number of young people not in education or employment, and by the cries of those submitting hundreds of applications and being turned down by AI in now dehumanised appointment processes. Our young people are more than data points, more than slaves to the algorithm and more than consumers. This kind of societal change is not inevitable. Technological change is not always progress. AI development needs the brake as well as the accelerator, as one of the founders of Anthropic said last night.
The UK can be, and needs to be, structured in such a way that there are meaningful pathways into good, creative employment and lifelong work as the foundation for human dignity, well-being and stable family life. We therefore need a much bigger, wider conversation on the societal impact of AI, drawing in the whole of our society. Governments need a richer, deeper understanding of what it means to be human, beyond creating a nation of consumers to a nation of citizens, drawing on the ancient paths of wisdom. That conversation will then be able to shape our economy for the benefit of families and society, not simply the profits of technology companies.
We need more investment, not less, in entry-level jobs, a rehumanising of the application and support process, more dialogues on the risks and dangers of working with algorithms, and some bigger thinking. A century ago, our nation responded to another new technology with the founding of the BBC—a radical step for the common good. We will need similar new institutions for the public benefit.
I sincerely hope for a response from the Minister to this debate which will demonstrate humility and acknowledge the crossroads, the jeopardy, and which will seek to broaden and deepen the conversations on the choices we must make.
Lord Johnson of Lainston (Con)
My Lords, I begin by declaring an interest. I am invested in AI companies, and I am hoping to go on the board of an AI company, because I think AI is the most amazing thing ever to happen to us. I cannot understand the long faces in this Chamber, the worry, the hand-wringing, about a technology that will change the world for the better. But it will be highly disruptive—I do not disagree with that. As we enter an age of abundance, I am pleased to say to the noble Lords opposite that it will become the end of capitalism. If resources can become effectively free, other things then become rationed. It could be leisure: if a billion people can go to Venice, obviously, how you access Venice will be determined by the capitalist principle of a market price. However, all the things we need to survive and thrive and enjoy ourselves will become free.
We are going to see massive disruption as a result— I do not disagree with the right reverend Prelate the Bishop of Oxford with regard to the future of work. While the artisanal classes will flourish as the cost of what they want comes down and the value of what they provide goes up, the white collar—mostly we noble Lords in this House—will be seriously disintermediated.
We are all saying that that is a bad thing, but it is not—it is a fabulous thing. Who wants to be an accountant, a lawyer, a consultant, a civil servant or even a doctor? Domain expertise will become irrelevant as technology takes over. We will not have to do these things. I do not belittle any noble Lord for having had a wonderful career, no doubt, in one of those occupations, and I do not want to sound flippant when I say that the end of those occupations will cause severe disruption to our body politic and society. But in the same way we do not say now, “Let’s get rid of our dishwasher or tumble dryer”, the fact is that the highest number of people in employment in 1900 were in domestic service, carrying water, heating it up, washing dishes and cleaning. No one wants to go back to that. We will be freed. We should look at this as the principle of freedom.
What worries me, though, is that we are not really thinking properly in government and in political parties about how this will affect society. That is why this debate, led by the most reverend Primate the Archbishop of Canterbury, is so important. If I think of the extraordinary revolutionary change between my life and my children’s, we really now understand the brick wall we are heading into. If we do not do a handbrake turn now, we will have serious repercussions.
If I think of my life, I was born, I went to school, I left university at 21, I got a job, I did the same job for 30 years, I bought a car, I got married, I bought a house, I had children; soon I will retire, and then I will die. That is not a particularly exciting prospect—I am afraid most noble Lords will share the same fate. But my 18-year-old daughter is not going to finish education—she is doing her A-levels at the moment. She has been educated entirely pointlessly. I wish her the best for her exams, but she is learning Edwardian, post-industrial service industry skills that are totally inappropriate for the future. She is not going to do the same job for 30 years; in the next two years, or even tomorrow—we are hearing how the disintermediation is happening already—she will do something completely different that we have not even imagined. She is not going to buy a house in the same way, her mobility will be different, she probably will not retire, and anyone born after 2005 might not die.
We are looking today at a government system that is entirely outdated, not just by a tweak—not just issues over energy or employment policy, or the disastrous consequences of poor Tory government or, worse, Labour government. We are looking at an entire societal change that we are simply not prepared for or thinking about. This is a crisis and emergency, not in terms of how we control AI but how we teach our citizens and educate ourselves to cope with an age of purposelessness, which is far more fundamental.
That is why this debate from the most reverend Primate the Archbishop of Canterbury is so fascinating. In fact, this is a wonderful thing, because we will be allowed to be more human. We go back to the ancient traditions of rhetoric, where we have grammar, logic and articulation. We have to learn information, but then we have to have the dialectic: we have to learn truth, which is what puzzles us—what is true in the information age?—and then, of course, how we articulate it to have voice. We can be more human.
I am excited for the Church of England. You can dust off all those pews. Those empty churches will be filled, because we will have so much time to actually have societal relationships, to think about how we work in our communities, to be who we want to be; we will not have to do those functions that took over our lives for the whole of the history of humanity.
So I ask the Government to be brave and bold about a vision. How will we deal with societies where you do not have the common purpose of the services sector, and where machines take the drudgery away from us? How do we teach ourselves to deal with these new luxuries? Without that debate, we will have a serious crisis, but if we master it, this country can take a lead in the AI revolution.
Baroness Gill (Lab)
My Lords, I express my sincere gratitude to the most reverend Primate the Archbishop of Canterbury for bringing this vital topic for debate today. The real impact of AI on our culture is not a theoretical threat to the future; it is a quiet, comfortable surrender of the human mind happening right now in our own living rooms.
I feel a sense of responsibility on this subject. Years ago, I sat in the European Parliament as a rapporteur for e-content, fighting to protect our unique cultural identity. Later, I worked for an IT company, specialising in data analytics. Looking back, we were building the foundational architecture of what we now call AI. In those days, I believed data analytics was a real force for good, and it was. We used data to target and reduce maternal mortality and the spread of AIDS, and to optimise global shipping routes to cut fuel consumption. So I am no Luddite. In another parliamentary report I authored, on elder care, I fiercely advocated for technology such as robotic surgery and smart home aids, to help our elders remain independent, dignified and safe in their homes.
However, we must recognise the fundamental difference between that era of targeted problem-solving and the sweeping algorithmic landscape of today. We are rattling blindly towards a world where we trade our core values for total convenience with Alexa, Google and Apple. Look at our families. We are witnessing the death of domestic conversation and connections. Go into any home today and you will see families who hardly speak to one another because every individual is entirely engrossed in their own isolated, personalised screen. Many have outsourced the mentorship and emotional development of our children.
More profoundly, our deepest spiritual values and religious traditions are under existential threat. Faith and culture rely on shared generational human connection, but an algorithm has no soul. By allowing automated systems to create our morality, we erode the very foundation of our shared values and ethical life.
Why are we letting this happen? No one, let alone a politician, wants to come across as a Luddite, so we walk on eggshells, terrified of appearing anti-progress. The consequence is a direct threat to our democracy and the security we took for granted for decades—that communal violence and disorder did not erupt instantaneously. We had a robust social fabric built on trust.
Look how fragile our society has become. We saw it in Southport, and in the last days we saw it in Southampton. A single piece of AI-amplified fake news, weaponised by algorithms optimised for outrage, shattered our societal consensus in a matter of hours. The damage travels further into politics, directly targeting our children. We are standing by as big tech’s “algorithms of outrage” systematically funnel young, isolated boys down toxic rabbit holes, funnelling them into the arms of mindless aggressors.
Combine that with the rise of harmful sites that actively encourage violence or aid and abet suicides among vulnerable young people, and this is where the role of the Government becomes critical. The UK has favoured an agile sector-by-sector approach, relying on regulators such as Ofcom. Agility is an advantage only if our regulators have teeth.
I ask the Minister to reflect on this. Can we truly expect voluntary codes of conduct to protect the society we have known? Will the Government move beyond mere guidelines and introduce a statutory opt-in right, legally compelling tech companies to let users turn off these toxic recommender algorithms entirely? We must draw a hard line at where automation stops aiding human capability and starts replacing human identity, as in superintelligent AI. Let us ensure that the digital architecture we guide continues to serve human wisdom, rather than automating humanity out of existence. We must fiercely reject an algorithm that replaces a family’s conversation, a community’s trust or citizens’ independent thought.
Years ago, I asked how to distribute digital information. Today, this House must ask how to protect the human soul within it.
Lord Shinkwin (Con) [V]
My Lords, I too congratulate the right reverend Primate on securing this important debate and on her thought-provoking speech. She and others have referred to the Pope’s encyclical on AI, which states:
“technology is never neutral, because it takes on the characteristics of those who devise, finance, regulate, and use it.”
I readily acknowledge, as others, including the noble Lord, Lord Ahmad of Wimbledon, have done, the potential benefits of AI to disabled people such as me. This autumn will see the launch of an exciting new initiative called Kerbcut, an AI tool to enable disabled people and those professionals who support them to find the right technology and support at the right time. I commend its founder and CEO, Hector Minto, formerly of Microsoft, for his passion to put his expertise at the service of others for the common good. But I doubt that that is what the Pope is alluding to when he speaks of the characteristics of those who devise, finance, regulate and use technology.
No one disputes that AI has placed immense power in the hands of unaccountable corporations and of individuals such as Elon Musk. As the right reverend Primate reminded us, power corrupts. Surely, at a time when the co-founder and CEO of Anthropic, the world’s most valuable start-up in AI, calls for a global freeze in AI development and warns that humans risk losing control of the technology, it is worth reflecting on what those characteristics to which the Pope alludes are, and how they could be exploited by a superintelligent, recursively self-improved AI ecosystem that speaks to itself, potentially bypassing us, as its creators, to our disastrous detriment.
We may laugh at “The Traitors” as a bit of light-hearted entertainment, but when we glorify the ability to deceive each other, should we really be surprised that an AI ecosystem that we have constructed deceives us, as apparently it is already showing signs of being able to do? In that situation, the concept of a kill switch is meaningless.
What other human characteristics might the AI ecosystem exploit? What about greed? Consider the Chinese genocide of Uyghur Muslims being perpetrated by the odious Communist Party regime, so odious that world leaders continue to queue up to ask for more business. I call it the four “G”s: greed gives the green light to genocide. We can be confident that Xi Jinping is already harnessing AI to commit genocide in Xinjiang province.
I conclude with this question for the Minister. Will the Government get behind the CEO of Anthropic and back his call for a halt in development so that we can retrofit ethics at the heart of AI before it is too late because, as the speaker before me said, the role of government here is critical?
Lord McNally (LD)
It is always a pleasure to follow the noble Lord, Lord Shinkwin, with his thoughtful and perceptive comments on our society. I have now been in the House of Lords for just over 30 years. When I accepted Paddy Ashdown’s invitation to take a nomination to this House in 1996, I genuinely thought that I would be coming here for two or three years before a new Labour Government would sweep away this House of patronage—and then, of course, of inheritance as well. For newcomers, I would not worry too much that this present Labour Government are going to rush, any more than did the 1997 Labour Government. But this does give me, like most people who have been here a long time, a great fondness for this House.
This morning has been a very good example of the House at its very best, as we wrap ourselves in the warm satisfaction of almost total agreement. AI is an ominous problem and it needs action. I have had the pleasure of serving on the Select Committee on Communications and Digital, skilfully chaired by the noble Baroness, Lady Keeley, and I have seen the expertise of that committee on view today from the noble Lord, Lord Tarassenko, and others.
The problem is that outside our cosy glow, there is a battle going on. The Minister is well aware of it, because she is seeking to deal with it womanly—I do not know what the feminine of “manfully” is. The problem is that this AI revolution is now under way, and I fear that there is a kind of paralysis at the heart of the Government at the moment. I see the pressures from the tech departments wanting to grab the benefits of AI that we have heard about and get them quickly forward. We have heard the concerns of many in the creative industries about the capacity of AI to dislodge and, as the noble Lord, Lord Johnson, explained, wipe away whole sections of employment.
My only worry is this. I had a very useful and enjoyable time at University College London, where I studied economic history—as well as student politics. My study of economic history reminds me that, as the noble Lord, Lord Johnson, said very frankly, in times of rapid technological change there will be mass casualties. In the past, Governments and societies have not always been able to handle those rapid changes within any kind of democratic process. That is what I fear we now face.
I was delighted when the most reverend Primate the Archbishop of Canterbury chose this issue for her first debate. When the Pope produced his first encyclical, almost on cue, it made me think that the links between Canterbury and Rome must be closer than we sometimes imagine. At the moment, the machinery of government does not have in place the kind of rapid response to deal with the problems that have been articulated so well today. I hope the Minister will give us some indication that the Government are putting in place a way for our democracy to handle these very difficult problems. If we do not do so, it will be those with the short, easy answers—the populists on both the left and the right—who get popular support for their solutions.
Baroness Prashar (CB)
My Lords, I, too, thank the most reverend Primate the Archbishop of Canterbury for securing this debate and for an extremely thoughtful introduction.
As we know, the central question before us is: how can we best harness the advantages and potential of AI without damaging society and humanity? Technology has always changed human behaviour, but AI is different because it reaches into areas once thought uniquely human: judgment, creativity, communication and even companionship. Human relationships are not built simply on the exchange of information. They are built on trust, empathy, understanding and shared experience, and they depend on our capacity to listen, to question and to challenge.
The misuse of AI destroys trust when people can no longer distinguish between what is real and what is artificially generated. Democratic debate is distorted when misinformation is produced at scale and targeted with unprecedented precision. Trusted information is a core part of the security infrastructure of a democratic society, and it must be a strategic priority for the Government.
Social cohesion frays when algorithms increasingly shape what we see, what we believe and whom we trust. As has rightly been referenced, the Pope recently stated that technology must remain a tool in the service of humanity, not its master; and wisdom is not as same as knowledge. One is reminded of Tennyson, who said:
“Knowledge comes, but wisdom lingers”.
For me, AI could also be called an ancient intelligence. It can process vast amounts of data, but it cannot exercise judgment. It cannot distinguish between what is technically possible and what is ethically right. As legislators, we have a responsibility to ensure that the development of AI is guided by principles that protect humanity and preserve accountability, transparency and democratic oversight, as well as being inclusive. It must not replace the human qualities upon which society is built.
We need proper governance to ensure that these technologies preserve the common good and shared prosperity. Governance is about not just rules, norms and procedures but incentives, eliminating barriers, fostering impact through investment and putting money into innovation to help navigate the challenges that arise from technology. The best governance systems will be those that can monitor the entire cycle of technology, infrastructure, algorithms and data. It is a myth that governance stifles innovation.
The Government’s plans for AI are rightly ambitious. The previous proposals, in 2023, which established the AI Safety Institute—it is now called the AI Security Institute—were geared towards mitigating future risks, but they are now missing from the current plan. The Government should not be reduced just to managing harms after they arise, which will narrow the choices going forward. Given AI’s potential, it is preferable to manage the risks early, rather than resorting to drastic measures later.
The AI Security Institute has made a useful contribution and was a good innovation. However, its work is retrospective, its scope was narrowed from safety to security, and it does not focus on societal harms such as misinformation. Voluntary compliance is no longer sufficient; what we need is a statutory framework. In yesterday’s debate, the noble Lord, Lord Holmes of Richmond, made a very good case for having a cross-sector, principles-based approach.
Going forward, there will also be a need for what I call AI literacy—equipping people for the new demands posed by AI by giving them the skills and qualities needed to use it intelligently, identifying and developing skills and talent for the future workforce, and offering opportunities for lifelong learning. Here, I support the call from the noble Lord, Lord Waldegrave, for higher preference being given to education in the humanities, which has been demoted in previous years. It would be useful to know what the Government are doing in this area, because I suspect they are lagging behind.
If we do not build the future for ourselves, AI will. We can take action now to show that progress rooted in ethics and the protection of humanity is not a constraint but a competitive edge and the foundation of a strong, democratic society.
Lord Taylor of Warwick (Non-Afl)
My Lords, I, too, thank the most reverend Primate the Archbishop of Canterbury for introducing this important debate. Through artificial intelligence, what was once considered science fiction has become science fact. Innovation and technology have brought solutions to people’s problems. Britain can be proud of the fact that it started the Industrial Revolution, but we are now in the fast-moving AI revolution, and the bait hides a hook.
In 1997, I had the pleasure of presenting the Bill that established Britain’s first ever comprehensive DNA database. However, without DNA regulation, we would not have experienced that positive DNA revolution. As a result, we became the first nation in the world to embrace this ground-breaking technology. Nearly 30 years later, the DNA database has positively transformed the effectiveness of medicine, forensic science and research science. Some 175 of the 195 nations in the world now utilise DNA profiles.
In the King’s Speech, there was no mention of an AI Bill. In January last year, the Government announced their intention to consult on proposed legislation to regulate AI and put the AI Security Institute on a statutory footing. Nearly 18 months later, the Government have not acted on these plans. I ask the Minister: why not?
Over half the adults and young people in the UK use artificial intelligence. This can produce human-like content and dialogue. However, we must not forget that each one of us has AI—it is called actual intelligence. There is a danger that increased use of AI for “friendship” could reduce incentives to maintain human relationships, with negative long-term social consequences. AI companions are always available and do not have their own needs, so people may find interacting with an AI friend easier than sustaining a human relationship.
For 10 years, I had the privilege of being vice-president of the British Board of Film Classification. We regulated films and video games from all over the world but particularly from Hollywood. Our remit was to protect children and other vulnerable sections of society. We saw how powerful films and video games can be in affecting emotions and ways in which the world is perceived. AI is having an increasing effect on these creative industries, including with films and video games. There must be AI guardrails. It is easier to build a strong child than to repair a broken man.
I am encouraged by the potential AI applications in healthcare. We are all living longer and the demands on the National Health Service are greater than ever. More money is being spent on the NHS, but simply spending more money is not the answer. It is like pouring water into a bath with the plug out. AI is creating new ways to treat illness and streamlining the management to make it more effective. However, there must be oversight from human doctors, nurses and other medical staff.
AI brings with it other challenges which are yet to be met. They include the impact on jobs and careers, the protection of copyright in the creative industries and the potential impact on the environment. AI models require large amounts of electricity, with increased carbon emissions. Each use of an AI tool needs a significant amount of energy. A ChatGPT query uses about five times more energy than a simple web search.
The irony is that the growth of AI technology could find its biggest obstacle in the lack of physical buildings to develop it in. Planning laws are available to people who oppose data centres in their locality. The energy consumption of data centres globally has increased greatly since 2022. It is expected that, by the end of this year, data centres will consume 1,050 terawatt hours—equal to the amount of the electricity consumption for Japan and Russia. Even the mining of raw materials needed to make the processors used for AI often involve harmful effects on the environment.
Some critics claim that AI machines may become more and more like humans. The danger may be that humans try to become more like machines, with a constant search for perfection and efficiency. It is the human soul that will always be unique. Beethoven created his wonderful ninth symphony to inspire the world at a time when he had become totally deaf and could not hear the music himself. Human beings, not machines, have the qualities of courage, faith, love, intuition, curiosity, imagination and creativity. They also experience sadness, failure, ill health and doubt. Yes, a machine can be manufactured with no visible imperfections or cracks, but in humans it is sometimes the cracks that let the light in.
Baroness Helic (Con)
My Lords, I am grateful to the most reverend Primate for securing this debate on a matter that goes to the heart of what kind of society we wish to be and what kind of humanity we wish to embody. Many of my points have already been made by noble Lords across the House. I shall none the less repeat some of them as they warrant repetition.
Artificial intelligence is no longer a distant prospect. It is reshaping our economy, our public life and how we relate to one another, at a speed that our institutions, our laws and our moral frameworks struggle to match. As other noble Lords have pointed out, there is much to welcome—advances in medicine, scientific discovery, productivity and the potential for better public services—but there is also much that should give us pause.
This is not, at its core, a debate about technology; it is a debate about power—who holds it, how it is exercised and whether human beings remain AI’s authors or become, by degrees, its subjects. The concentration of data and computational capacity in the hands of a small number of corporations is not a technical detail; it is a political fact, with consequences for sovereignty, democratic accountability and the rights of ordinary citizens, who have little influence over how these systems are designed or deployed. The question before us is not whether artificial intelligence will continue to advance. It will. The question is whether democratic societies will exercise meaningful control over how it shapes our lives or whether we drift into a future increasingly governed by systems that few understand and fewer still can challenge.
This question bears particular weight on the young. As has already been said, we hear from students and graduates who are uncertain about the future for which they have prepared. Job offers are being withdrawn, entry-level roles are disappearing and career pathways are becoming less clear. What assessment have the Government made of AI’s impact on graduate employment and social mobility? How will they ensure that the gains from this technology are shared rather than concentrated? Progress that excludes a generation is not progress at all.
It is sometimes argued that innovation should simply be allowed to take its course and that markets will determine the best outcome. However, markets do not safeguard human dignity, uphold democratic values or distribute benefits fairly. These are matters of governance and political judgment. To abdicate that responsibility in the name of competitiveness would be a serious failure of duty.
Our democratic life is also at risk, in ways that we are only beginning to understand. AI systems increasingly shape what people read, hear and believe. They influence public debate, political discourse and the formation of opinion. What safeguards do the Government believe are necessary to protect democratic processes from AI-enabled manipulation and disinformation? How will citizens know that the information they receive has not been generated, altered or amplified by artificial intelligence?
Nowhere are the stakes higher than in warfare. Artificial intelligence is already embedded in surveillance and military decision-making. Systems are now being developed that can select and engage targets with diminishing levels of human intervention. The question of who is accountable when such a system fails and leads to the loss of civilian life—for instance, the terrible recent US bombing of an Iranian primary school for girls—has no satisfactory answer. Existing legal frameworks were written on the assumption that human beings would exercise judgment, restraint and moral responsibility and that such matters would not be left to machines.
What is the Government’s position, therefore, on maintaining meaningful human control over the use of lethal force? Do they support international efforts to regulate or prohibit fully autonomous weapons systems? A phrase increasingly heard in defence and arms-control circles is the “Oppenheimer moment”. It reflects a growing recognition that autonomous weapons may represent a technological threshold comparable in significance with the advent of nuclear weapons. Unlike nuclear weapons, however, autonomous systems are comparatively cheap, scalable and accessible. The barriers to proliferation are low while the potential for misuse is considerable.
Yet we are not witnessing the kind of sustained international negotiation that the gravity of this challenge demands. It is shocking that there is still no prospect of any comprehensive international framework governing autonomous weapons. What leadership is the United Kingdom providing in building one? This matters morally as well as strategically. One of the most troubling features of autonomous weapons is the distance that they place between the act of killing and any human sense of responsibility. That distance does not lessen the suffering of victims but merely removes the human conscience that might otherwise have prevented it.
Artificial intelligence must serve humanity, not the other way around. Its development can be justified only if it remains anchored in human dignity, democratic accountability and moral responsibility. The choices that we make now will shape not only the technologies of the future but the character of the society which that future inherits.
Lord Brooke of Alverthorpe (Lab)
I too am most grateful to the most reverend Primate for selecting this important topic, for her wide-ranging and stimulating speech and for proposing the three questions that we should address. On her third, the truth, it is worth remembering that there was life on this planet before humanity. Notwithstanding our desire that humanity should not be replaced or surpassed, it could be, if it is God's will.
We should pay particular attention to the fact that there are engineers who are seeking to implant into the AI technology itself, as it is developing, those emotions, feelings and characteristics of human beings that the noble Lord, Lord Ahmad, referred to. It is not going to be easy to stop this, slow it or reverse it, but of course we should try. At a minimum, we should endeavour to guide the changes—we should at least be trying to get wisdom and goodness into AI.
We should make greater efforts to be in control of our destiny. We rely too much on the USA’s big tech companies. I have long been an advocate of seeking to create our own UK sovereign AI entity on not dissimilar lines to what was mentioned, like the way we created the BBC in the 1930s. The first Oral Question we had this week was on this topic. In particular, my colleague, the noble Baroness, Lady Berger, pointed out that:
“France, Germany, South Korea, Switzerland, Singapore, Canada and Ukraine are all investing in”—[Official Report, 1/6/26; col. 628.]
this. She overlooked to mention that India is launching its own and will be, I believe, able to compete with China and the big American tech companies. Perhaps, if we cannot do it ourselves, we at least ought to be thinking of trying to establish a partnership with it. But before that, why is the UK not investing in its own model? That is what I ask the Minister.
The noble Lord, Lord Walney, earlier this week pressed a similar question. My noble friend the Minister replied that she could not give a definitive opinion. I trust that she has taken some further soundings since we had that debate earlier in the week. We also had a very important debate yesterday in the moses Room, on governance and regulation. I look forward to her responding, I hope positively, to the desire which has been expressed by an increasing number that, notwithstanding the cost, we should seek to have greater independence in this field.
We can look to try to develop a public/private partnership model, which would be wider than the ones that we have used in the past, with more people involved with the public; in particular, we should extend the private over a wider area. We should try to get parents and other people who are keen to see a UK sovereign AI entity operating to invest into it. I pose the question to the most reverend Primate of whether the Church itself might consider investing in a UK AI entity, so that we have control over the machine and over what our children should be encountering rather than leaving it to others to do so.
Similarly, given that the Pope has made such an important intervention, can we look forward to the Churches coming together, not just within the Christian faith but over the broader faith throughout the world, so that we can see a wider response taking place? That is what is going to be needed: a response in the way that we have endeavoured to do with climate change. Regrettably, somebody always breaks ranks. This is the problem that we encounter because of our humanity. But we must try. We must have faith, and we must look to see that there is a better world before us, notwithstanding the challenges which come on a daily basis from new technology.
Lord Johnson of Marylebone (Con)
My Lords, I echo the gratitude to the most reverend Primate for securing this important debate.
AI is a technology that changes how human beings relate to knowledge and authority and therefore to one another. For much of the modern era, authority has derived in part from the possession of knowledge and expertise. In the age of AI, both are commoditised. Who will control supply of that commodity?
Earlier this year, chillingly to my mind, Sam Altman, CEO of OpenAI, one of the AI companies which is about to list on the US stock market with a trillion dollar-plus valuation, said he predicted
“a future where intelligence is a utility, like electricity or water, and people buy it from us on a meter”.
Indeed, all over the world, students already are interrogating AI systems about constitutional law, quantum physics or Shakespeare, using their subscriptions to OpenAI and other such services to do so, and they are receiving answers that appear as sophisticated as any given by experts.
Our education system, especially our universities with their vital but hard-pressed humanities departments, are grappling with this momentous development. AI challenges not just the role of the professor as the guardian and gatekeeper of disciplinary knowledge and the way we assess learning but the deeper purpose of education itself. Ultimately, the value of higher education cannot simply be in the transmission of knowledge. It must be in the purposeful cultivation, as other noble Lords have said today, of human judgment—the ability to detect disinformation and recognise a beguiling but false argument, and to disagree with each other in a civilised fashion.
Large language models make the production of plausible-looking answers trivially easy, and for many students, sadly, almost impossible to resist. But LLMs do not know and do not care if judgment has been formed. They are built to produce fluent outputs, not to nurture epistemic responsibility and actual intellectual growth. The emerging evidence is not that AI inevitably destroys critical thinking. Used well, hybrid intelligence can expand our capabilities, not just substitute them; but used badly, as I fear is increasingly the risk, as an answer machine rather than as a thinking partner, it encourages a damaging form of cognitive offloading that really does weaken our capacity for critical thinking and judgment.
Universities as institutions that develop critical thinking perhaps more than most—or at least they are meant to—matter more than ever in the age of AI, not less, but we should be honest: the public are increasingly sceptical that they can actually perform these functions. Earlier this week, in polling from the British Social Attitudes organisation, we saw that scepticism about the value of a degree is now at a record high, with 34% of people polled in England saying that higher education is not worth the time and money it takes, up from 14% in 2005. Many factors might be at play here, but I am sure AI is among them.
In the age of AI, universities, as critical institutions in our society, have to regain public trust not by asserting their authority but by demonstrating that they can uphold standards, protect intellectual independence and help us form judgment in ways machines cannot. The task before us is not to resist artificial intelligence; it is to ensure that, in embracing it, we do not fall prey to the temptation of cognitive offloading to the extent that we weaken the capacity to develop judgment of our own.
Lord Hastings of Scarisbrick (CB)
My Lords, in thanking the most reverend Primate the Archbishop for introducing this debate with such excellence, I refer to another spiritual icon of our age, from the cartoon Peanuts. In a wonderful cartoon, in which a character sits fishing by the side of a river, he says:
“I don’t want a city on Mars. I don’t want AI in every app. I don’t want data centers in space. I want clean water. I want affordable health care. I want bees to survive”.
Most of our lives are about basic things, not extremes, and we need to treasure the basic things before we get enveloped in the externalities.
Albert Einstein in his final years warned: “I fear the day that technology will surpass our human interaction. The world will have a generation of idiots”. It is up to all of us who are involved in education and universities— I chair one, am chancellor of another and professor of leadership at a third. My students, 20,000 of them in the US and here in the UK, treat AI as a utility. It is a bit like McDonald’s on a daily basis: it is not a healthy menu and it does not give choices to avoid extreme addiction.
Students and those whom I see on a daily basis, who are the minority in our society, have the ability for careful, analytical and adaptive thinking—well, many of them do. They can also ask complex questions. They can learn to value humanities as much as they value science. But the majority of our society does not have those inbuilt or trained abilities. Therefore, I launch at the Minister a request that we start to treat the understanding of AI and the technology linked to it, good as it is in so many wonderful ways, as an emerging and serious addiction. In the same way that health departments recognised that it was important to have public campaigns to lead us to understand the addictive nature of tobacco and alcohol, we must now understand the addictive nature of AI and abusive technologies. We need to be warned about them. It is a public duty.
I spent part of this week listening to a group of recent ex-offenders online and in person—10 on Saturday and six on Tuesday night on Zoom. They were being asked to take part in an Oxford University investigation on the role of AI in preventing repeat offending. It was interesting to listen. I said nothing, just listened, as they gave their very careful thoughts, having between them over 126 years of incarcerated time in our prisons. In reflecting on the opportunities of AI—yes, it is easy to come out of prison and immediately there is all your data on the system; a driving licence, an NHS number, a national insurance number and, of course, a higher tax bill—they said three things that were, rightly, warnings.
First, they had spent literally decades without technology and were intrinsically suspicious of how data transfer and drift will lead to manipulation of their lives. That is relevant not just to those who are ex-offenders but to anyone who carries risk in the nature of their past. That risk can be transferred by data systems that are interested not in respecting human dignity but just in getting the job done.
Secondly, they warned against assumptions that AI-driven health knowledge was the right knowledge, given that many of them have profiles where the difficult issues that led them to prison are assumed to be their continuing reality—in other words, once depressed, maybe even once highly negative, they remain so. Therefore, health solutions given by AI doctors since coming out of prison have always encouraged them towards anti-depressant systems, even when they do not feel that way. Thirdly, they highlighted the ways in which AI pushes the possibilities of suicidal thinking. Some have felt that pressure.
They rightly warn us that we need to prepare ourselves to be better and highly educated, very cautious and risk-centric and to treat AI for what it is: a benefit that, a bit like too much McDonald’s, destroys the whole soul, let alone the body.
Baroness Owen of Alderley Edge (Con)
My Lords, I declare an interest as a guest of Google at its AI policy conference, as a member of the advisory board for Arãya Ventures, a venture capital company that invests in AI, and in receiving pro bono legal advice from Mishcon de Reya on image-based sexual abuse. I am grateful to the most reverend Primate the Archbishop of Canterbury for tabling this debate and I congratulate her on taking up that historic role and being the first woman to hold it.
Artificial intelligence is the defining technology of our time. It has the power to cure disease, accelerate scientific discovery and transform our economy. Yet it also has the power to deceive, manipulate and even take life through autonomous weapons systems. Your Lordships’ House has frequently proposed legislation to regulate and mitigate against its worst impacts and it is vital that we continue this important work, especially in respect of the appalling cases raised by the noble Baroness, Lady Kidron, of chatbots coercing children to self-harm and suicide.
Therefore, it is right that we consider the impact of AI on humans. But I cannot help but wonder whether we have got the question the wrong way round. Is artificial intelligence not at its core an extension of those who have trained and developed it? Does it not hold up a mirror to where our values have already slipped?
Noble Lords will know that I am passionate about preventing people from using AI to create non-consensual sexually explicit images. I firmly believe that the driving force behind that particular application of AI technology is ingrained misogyny within our society, which has utilised AI as a powerful new frontier. These applications were not accidents; the AI models were trained or fine-tuned with a purpose—to violate a woman’s consent. The root cause lies not in the AI technology but in those who trained or fine-tuned it. Therefore, perhaps the question before us should be the impact of humans on artificial intelligence, rather than the other way round.
In his encyclical Magnifica Humanitas last week, the Pope warned that those who control AI will impose their own moral vision, which will become the invisible infrastructure of these systems. A more moral AI is not enough if that morality is determined by a few. I wholly agree that the development of AI should not be left to a handful of individuals in Silicon Valley, but I have deep concerns that by seeking to slow its progress, we simply count ourselves out of the race and cede the power to shape it to those whose values we may not share.
In an interview with the Financial Times this week, Jack Clark, a co-founder of Anthropic, said:
“If it were possible to elegantly slow the development of this technology to give us more time … to deal with this, that would likely be a good thing”.
But he acknowledged that slowing down would be immensely difficult given the commercial and geopolitical rivalries with China.
We must always keep in mind what this technology can achieve when utilised for the benefit of society. One need only take the example of AlphaFold by Google DeepMind, which has predicted 3D protein structures with an accuracy that would have taken conventional research many years to achieve, ultimately winning the Nobel Prize in Chemistry for its chief developers. That work is now being utilised by millions of scientists around the world, accelerating the development of treatments for diseases that have long defeated us.
We have seen AI used to identify vulnerabilities in critical infrastructure and to give a voice to those who have lost the ability to speak. Whether we like it or not, if we as a country do not embrace this technology, this technology will leave us behind. It is our responsibility to ensure that this Government do everything in their power to make the UK a genuine leader in AI, equipping our society with the skills and adaptability to meet what is coming and ensuring that our laws and values are woven into its development from the start. Without partaking, we will have no power to shape it for better or for worse. Only by being present in this race can we ensure that we do not cede the ability to shape our own future to those whose values we do not share.
Lord Griffiths of Burry Port (Lab)
My Lords, AI crossed the threshold of my understanding and experience much earlier than for any other speaker in this debate thus far. When I left for an underdeveloped country and was given some responsibility for a rural rehabilitation project, I found that it was all centred on AI— but then, of course, it meant artificial insemination. The letters have remained firmly in my mind, and the improvements to animal husbandry and all the rest of it that I was to learn about were a direct consequence of that particular branch of scientific development.
I add my voice to all the others who thanked my dear friend the most reverend Primate the Archbishop of Canterbury for bringing this debate to the Floor of the House today. She had such an adroit sense of timing that it follows not only the Vatican’s encyclical, but also Anne Keast-Butler, head of GCHQ, speaking a day later at Bletchley Park about the place of modern technology in our defence and security sectors. A day after that, Alan Milburn’s Young People and Work was published, to the consternation of the entire population as to the likely consequences of artificial intelligence in the world of work.
While it is true, as an earlier speaker said, that over time what happens in the next few years will have been assimilated and we will have grown skills and experience that will enable us, at that stage, to make progress, it remains a fact that people coming out of university now, with enormous debts, will find work very difficult to find. The capacity to bring a debate to this House and to stimulate, as it seemed, all these other responses in other fields is a great skill that only an Archbishop of Canterbury can pull off.
For three years I was a member of the Communications and Digital Select Committee with others in this Chamber. There is an aspect of the work that we covered there that has not been mentioned in the debate thus far. Underlying any remarks I make is a total and ready understanding of the need to accept and honour the work of scientists in bringing all these developments and improvements to our knowledge and for our use. I do not want any of that to enter my speech, but I want it to be understood that I cannot think of the world without those developments happening. Even in our report on large language models and generative AI, which was very extensive and saw many witnesses interrogated and in which we set ourselves the task of concentrating on the positive things that AI will contribute to the well-being—economic and otherwise—of the country, we still had a pretty loaded chapter on risk because we recognised that too. It would be very foolish for us to fall into the position where we are so keen to see and acknowledge the developments, and there has been ample evidence of that today, that we fail to see and take measures against the encroaching risks that come with the package.
While I was on that committee, we did another report on a category that has not been mentioned, digital exclusion. So many people in this country have no access, or no real access, to digital life. The report said there were 1.7 million such people. I cannot believe the figure is any better now, a year or two after that report was published. We went to hubs to see how people with no skills, no money and no stake in the economic development of our country are sharing information, facilitated by people who give their time to help to develop these things so that people have access to all the things they need access for: their rent, their welfare, their bank and all the rest of it. We should not be unaware of the fact that the other side of the progress that we expect is the creation of an underclass.
Also, in the period following the most reverend Primate the Archbishop—Oh! I am past my time. A verbose Welshman cannot possibly contribute to continuing this debate any longer, but let us remember the poor and the excluded. That is enough.
The Lord Bishop of Leicester
My Lords, in the past, if you wanted to persuade people to think badly of others, you were limited by two things: the number of people you personally knew and the number of conversations you could physically have. Social media removed the second of those limits, letting one person reach millions at once. Artificial intelligence is now loosening the first. It allows one person to produce vast quantities of content of increasingly high quality. The frictions that once limited the spread of contempt have disappeared. We should not, then, be surprised that the fabric of our society is being torn.
For society to function, we need a broadly common understanding of the world and what is happening in it. Democracy is about disagreements over what to do about the opportunities and challenges we face, but for that disagreement to be constructive, we must all be able to access the bare facts: what is happening, who is involved and who is affected? Generative AI throws all this into question. Anyone, anywhere can now produce an image of an event that has never occurred or a video of a public figure saying something they never said. I really do mean anyone, anywhere. The BBC recently reported that accounts producing AI-generated anti-immigration content that appeared to be British were in fact run from east Asia, the Gulf and the United States.
Even as the number and reach of deepfakes continue to grow, there is also the fact that AI allows people for the first time to visualise abstractions on demand. A fear or suspicion that once lived only in the imagination can now be rendered as an apparent photograph in seconds and shared to incite or confirm the same fear in others. This matters because human beings have always been moved more powerfully by images than by arguments. Importantly, what people see shapes how they act in the encounters of ordinary life and at the ballot box.
We know from Allport’s contact theory that what most helps people to let go of prejudicial abstractions is interpersonal encounter, particularly with a common purpose in mind. Here AI poses a further threat as a growing number of people are turning to AI companions for friendship, romantic intimacy, therapy or spiritual guidance. Systems are always there, always ready to listen. They never have a bad day. They have no ego or agenda of their own, apparently. They are in a sense the perfect partner, but that is the problem. As others have already said, real human relationships are difficult. They require us to tolerate frustration, to forgive, to be forgiven, to encounter a mind that is genuinely not our own. These are the muscles of social life.
I have two specific proposals to help ensure that we are intentional in the way AI develops and works for the good of relationships and our shared life in this nation. First, I suggest that we must require social media platforms to change their structural incentives so that they algorithmically deprioritise content damaging to public debate. As Frances Haugen, the former Facebook product manager, put it:
“Anger and hate is the easiest way to grow on Facebook”.
That should not be so. Secondly, following the EU, we should mandate a crisis protocol, a set of obligations that come into force when a platform’s content begins to threaten public order or social cohesion in a measurable way, as it did in Leicester in 2022, when false claims about attacks on Muslims and Hindus fuelled unrest.
In this parliamentary Session we have Bills before us concerning AI and others on extremism and state threats, but I dare to say that they barely scratch the surface of what I have described. They do not touch the structural incentives behind inflammatory content, and nor do they support positive connection across difference. This is what the Church of England, like so many in civil society organisations, wants to work towards, but we need the legislative and regulatory framework to be able to have the maximum impact.
Lord Holmes of Richmond (Con)
My Lords, I declare my relevant interests as adviser to the Crown Estate, Endava PLC and Simmons & Simmons LLP.
I congratulate the most reverend Primate not only on securing this debate but, much more importantly, on the short journey she has taken from Saint Paul’s to Lambeth Palace and, through that, the magnificent distance she has travelled for all those who will follow and for the entire global communion.
In considering that question of travel, it is worth considering how far AI has travelled. A good place to start is with the good book. Some decades ago, a basic ML sought to translate the Bible from English into Russian. For Matthew’s sage comment,
“the spirit is willing, but the flesh is weak”,
the translation came up with “the vodka is good, but the meat is rotten”. We have all experienced such meals from time to time.
It is fair to say that AI has travelled more than a cosmos since then, and it is for that reason—putting the human and the social context on this—I wrote a report last year, 8 Realities, 8 Billion Reasons to Regulate, taking eight human experiences and archetyping them for general effect. Let us take the school, the workplace and the therapy setting: just three.
In school, a smartphone could be such a positive deliverer of personalised learning, not least for those who find themselves all too often at the sharp end: those with special educational needs or disabilities. Yet, because of the platforms on those smartphones, we are on the brink—understandably and desperately unfortunately—of a ban. Those smartphones could still, potentially but not inevitably, be such a powerful classroom assistant for all in education.
Then there is the job seeker who is constantly rejected, not making the shortlist and not even knowing that AI is often behind sometimes biased decisions. Even if she or he knew that, they have no right or place to seek redress.
In the therapy setting, there is such positive potential from AI, yet we all know the horrific stories, many of them currently in the courts. Even the changes that have been made do not solve this issue. They offer management and legalistic solutions that simultaneously drop people mid-conversation and are similarly damaging to the most vulnerable in our societies. In short, how do we consider the outcome of hundreds of millions daily seeking online therapy and getting relational cues from bots that have no relational capacity? None of us knows the outcomes.
Ultimately, as ever, this is a question of belief: credo. Can we believe in the developers, deployers, designers and sellers of these systems? Should we believe in the business models that are built around them? Can we believe that our data—sometimes our most sensitive personal data—will be treated with dignity, respect, privacy and lack of bias?
Credo, credo, credo; we need more than good intentions. We know all too well what they have the potential to pave. We need urgent government action to bring forward principles-based, outcomes-focused, inputs-understood, right-sized, cross-sector legislation and attendant regulation. We need international connection and collaboration, working on international standards and agreements, all imbued with human values. We need to see the convening of public engagement the like of which we have never seen across our United Kingdom, so that everybody can express their fears and their hopes and see how they can share in the prosperity and the possibilities of these technologies.
Without this and more from government, I am afraid that government efforts with AI in the UK are “credo, not so much”. We find ourselves not in a position to optimise. We find ourselves, I am afraid, more like flies to little boys. We are flies to the bots. They kill us because they are bored.
Baroness Stuart of Edgbaston (CB)
My Lords, in the last four years, in the role of the First Civil Service Commissioner, I have increasingly seen AI being used in all stages of recruitment. The Civil Service uses AI and so does the commission, but we have always been very clear that AI is supposed to support the human decision-maker, rather than to remove the human decision-maker.
The Cabinet Manual, which we were told earlier this week would be updated, does not make any reference to AI, but it does say:
“Ministers have a duty to give fair consideration and due weight to informed and impartial advice from civil servants, as well as to other considerations and advice in reaching policy decisions”.
The core value of the Civil Service is objectivity, and basing your advice and decisions on rigorous analysis of the evidence. But where does that advice come from? What is the basis? What is the information we are using? AI is increasingly giving us a vast number of sources and data, but at some stage someone has to make a decision and, more to the point, someone has to use their judgment.
I then thought that I would use AI to ask about how to make judgments. It gave me two spellings: without the e, it is a court judgment, and the one with the e in it, which is slightly more nuanced, it told me was rather old-fashioned and falling out of use. I thought, “It may be old-fashioned, but I still quite like the nuance of those two”. I went back and thought, “Right, what do I do when I use judgment?” The advice came back with a whole checklist on when I use my judgment and examples of how you use your judgment, whether it is as a customer service manager or whether you decide to stop the car because the rain is too heavy.
But then I thought, “What will it say if I say that I am bad at making judgments?” It came back and said, “Would you like some help with self-assessment for professional applications?” I said, “Yes. What do I say in my self-assessment when I’m bad at making judgments?” It came back and said, “Don’t use the word ‘bad’. Say you are actively working to refine your decision-making processes”. I said, “But what if I always get it wrong?” It said, “Well, show how you learn from mistakes”. I thought, “How far do I have to go before the answer comes back, saying, ‘Well, maybe you’re not the right person for this job’?” It never did.
With that positivity bias, on the one hand, you could say, “Does it matter?” I would say “Yes, it does”, because all these things have underlying assumptions. In this example, they are very easy to spot, but with a great many others it is much more difficult. It gives you the strange belief that all you need is more information and you can resolve whatever question is coming your way.
I cast my mind back to many a Friday in the last Session which we spent on the assisted dying Bill. If I were to ask, “Would a society in which assisted dying is normalised be better because it has enhanced personal autonomy and reduced suffering, or worse, because it has devalued lives that others have judged are not worth living?”, I am afraid that AI could not help with that. We still have to develop the ability to make judgments. Technology is a tool, and we should always use it as a tool. Those who look to AI as a salvation, as the planetary superintelligence that echoes the absent god of monotheism, will be disappointed. But the sociopolitical shock of our workforce and the realisation that human intelligence is losing its scarcity value is a real one, and it is coming. The knowledge class will experience what the agricultural labourers experienced during the Industrial Revolution. The Luddites may have tried to smash the spinning jennies, but that never resolved the problem.
However, there is something that we can do, and something we can prevent, and it is our individual responsibility to do so. It is that human redundancy does not happen in personal relationships and it does not happen in our perceptions of reality. It is our own sense and responsibility to be clear that we do not end up saying, “Alexa, it’s over to you, you’re in charge now”. Ultimately, we remain in charge of our decision-making. We have a responsibility in our decision-making, and we have a responsibility to the next generation to allow them to make their judgments, to learn from their errors and to actually realise that moral judgment is what makes us human.
Lord Kamall (Con)
My Lords, I too thank the most reverend Primate for securing this important debate. While the speech outlined some of the more existential questions, I will try to build on the most reverend Primate’s comments on healthcare. It is also a pleasure to follow the noble Baroness, Lady Stuart, and I want to build on some of the comments she made about judgment.
In considering the demographic challenge of an ageing population, the increased demand on health and social care and the pressure on the public purse, the recently departed Secretary of State for Health and Social Care suggested that, instead of just spending more money and recruiting more people, the health system should use more AI. When we speak about AI, much of what we call AI today is really machine learning, trained on huge amounts of data to reveal hidden patterns, make predictions and learn over time.
We have the wonderful situation where, when a patient has an ocular scan, not only can the optometrist check the condition of the eye but, thanks to machine learning trained on huge medical data sets, they are able to identify whether patients also have other conditions, such as high blood pressure. So you can see why many are enthusiastic about the potential of AI in medical diagnosis but also in helping clinicians to make more informed decisions. However, taking that a step further and automating decision-making should give us pause for thought. Going even further, when we allow AI systems to rewrite their own algorithms, some see that as a step too far, with fears of machines enslaving humans and ruling the world.
Some of that AI is actually with us now. In a documentary about the use of AI by Ocado—you can see what an interesting guy I am when I watch documentaries like that—the manager explained that, while humans wrote the original algorithm, the system itself rewrites the algorithm to improve the efficiency of preparing the crates for delivery. He admitted that he no longer understood or knew what the algorithm was. For some, that will sound scary, but so far no Ocado robotic pickers have broken out of the warehouse, rampaged through the nearest town and left a trail of destruction.
With regard to medical applications in other areas, especially the military, there are concerns about fully automated processes, as a number of noble Lords have said, but even here it is not always clear cut. Consider AI-driven missile systems. While some have a human in the loop—that is, the AI identifies a legitimate target for a missile strike but a human operator still has to press the button—what happens when, effectively, the human operator says, “Actually, that system gets it right most of the time”, and just automatically presses that button? The same thing happens when we click the button for cookies—we just automatically click. That is a process of self-automation. Now imagine that in the healthcare system.
Interestingly, as an aside, when an AI algorithm was blamed for the recent missile strike that tragically murdered 180 schoolchildren in Iran, as mentioned by my noble friend Lady Helic, it turned out that the US military had not updated the data on that building, demonstrating that AI is not only only as good as the algorithm but only as good as the data it is trained on.
The other concern is that, while we see more use of AI in tech and commerce, the same systems may not always work in healthcare. I shall illustrate with a couple of examples. A few years ago, I arrived at an airport and scanned my boarding pass to get to the gate, but when I went to board the plane, the system was not working. I asked easyJet staff about it, and I was gaslit by many; over the next hour I had meeting after frustrating meeting, trying to find someone who would help me. The next day they admitted to me that a flag had gone off in the system—as if that explained everything.
Another example is when I applied for a Monzo bank account. I got a message saying, “We will process your application within 48 hours”, but a week later I had heard nothing. I chased them up, and a week later I got a message saying, “We’ve decided to reject your application, and by law we don’t have to tell you why”. The point here is that, while those companies can get away with that because there is competition and choice, imagine that happening in healthcare. You turn up for your operation, to be told, “I’m sorry, we don’t have to tell you why you’ve been declined for your operation, but you can’t get in”. So, while we should be excited about the huge potential of AI for medical research and diagnosis, when it comes to combined AI and automation for delivering health and care services, of course let us continue to innovate, but let us do so with caution and humanity.
Lord Frost (Non-Afl)
My Lords, I thank the most reverend Primate for her excellent opening speech in this debate, which admirably set out the issues that I think ought to be concerning us.
AI is undoubtedly going to bring us many good things, but it is a technology different in nature from any we have had before, perhaps even since the invention of writing, or at least printing. It has an affective quality that no technology has ever had before it. We have already stopped noticing the fundamental novelty, and indeed the oddness, of being able to talk to a machine in normal human language and get a reply. We have never been able to do that before, and I worry a bit about the consequences.
I am not personally concerned that all this is happening because AI is in some way genuinely intelligent or even sentient. I do not believe that humans are capable of creating such a thing. Some, of course, do think that, and Chris Olah, a founder of Anthropic, said in front of the Pope last week that
“we keep finding things that are mysterious, even unsettling. We find structures that mirror results from human neuroscience … We find internal states that functionally mirror joy, satisfaction, fear, grief, and unease”.
The problem with looking at things in this way is that confuses what is intelligible with what is intelligent.
The most reverend Primate the Archbishop mentioned CS Lewis earlier and, in what I am about to say, I draw on his great essay, Transposition, which is, as so often, a remarkable anticipation of current realities. A materialist looks at the human brain and thinks that brain activity and neuroscience can, in principle, explain everything. I look at the same phenomena and see higher realities, consciousness, the divine spark and things we do not fully understand—a fuller reality simply reflected in the lower reality of brain material, and showing up in neuroscience as that activity. If Chris Olah thinks that what he sees in AI mirrors human capacity in some way, it is because he looks at that lower activity in his AI model, notes that there is something similar in humans and assumes that they are the same thing. They are not. We have never shown that neuroscience can create consciousness; I do not believe we ever will. A mind cannot exist in a human-created machine. AI mimics intelligence but it is not, in itself, intelligent, as the Pope set out in Magnifica Humanitas.
My worry is that this leaves us with another kind of problem. That worry is set out in a very interesting recent essay in Nature by Mustafa Suleyman, now chief executive at Microsoft AI. He worries that people are starting to see AI as sentient even when it is not. As he puts it:
“When a system perfectly mimics intentionality and empathy, the human brain projects an inner life into it … We are hurtling into this era largely unprepared for the psychological fallout”.
I must say that I sympathise. I cannot think it good if large numbers of people think that their AI loves them. It is something anti-human, fundamentally “not normal”. It is especially worrying in a world where many people take a utilitarian view of humanity and believe human beings are nothing special and are somehow just computers embedded in a body.
Where I differ from those who have expressed similar concerns today is that I am sceptical about regulation as the way to solve these problems. Some noble Lords seem to have a, frankly, fantastical belief in the powers of the British state to control society, shape AI development and influence what is happening around the world. Even if we could regulate ourselves, I do not know that the Chinese Government, to take just one example, would do the same. There are things, however, that we can do as people. As a society, we should try to show as little social tolerance as we can for this idea of treating AI as a person, a disembodied adviser or some sort of daemon from one of Philip Pullman’s books. For the end of that road is moral and legal rights for machines, which would be a fundamentally destabilising and dangerous development. We cannot control the technology, but we can control our reaction to it.
We are not going to solve the problems presented by AI by seeking refuge in more and more government control; we will do it by maintaining our own confidence as people in the special nature of humanity and our fundamental difference from the technology, and hence our ability to master it and not be pulled in by the mesmerising quality of the machine. I must say that, as I believe myself, the established Abrahamic religions, with their integral concepts of the special divine nature of humanity, seem most likely to provide us with the concepts and moral strength to achieve this.
Fundamentally, to conclude, this is not a problem Governments can solve. It is a problem for society, for individuals and, therefore, for each one of us in what we say, what we believe and how we think.
Lord Watson of Invergowrie (Lab)
My Lords, the powerful opening speech by the most reverend Primate contained much wisdom. For me, however, perhaps the main takeaway will be a call for AI developers to answer the fundamental question: where are we going? The potential impact of artificial intelligence on human relationships and society is vast and far-reaching; it touches nearly every aspect of our lives, from how we work and learn to how we connect with one another.
On the potential impact of AI in society, several noble Lords have referenced Geoffrey Hinton, the godfather of AI. He has estimated that there is a 10% to 20% likelihood of AI leading to human extinction within a few decades. Earlier this year, a prominent safety expert, Mrinank Sharma, quit his job at Anthropic with the warning:
“The world is in peril”.
These are people who know what they are talking about when it comes to AI, and if they are worried, should not we lesser beings be worried as well?
The reason that experts are so concerned lies in how modern AI systems are developed: in a very real sense, they are grown rather than built. Systems are being created whose behaviour cannot be reliably anticipated, thereby producing actions humans may not be able to control. This challenge becomes even more serious as many AI companies are now openly pursuing the creation of superintelligent AI. These are systems that eclipse human performance across all cognitive tasks. If a system more intelligent than any human were to act in ways at odds with human values—which is very likely—the consequences could be catastrophic. The problem of maintaining control over AI systems is not theoretical; it is already taking shape.
One of the major effects AI will have on humans is economic displacement. As AI systems become increasingly capable, the risk that human labour will be replaced on a significant scale grows ever more real. Jobs that once required years of training and experience may become automated in a matter of months. Entire industries could be transformed, leaving millions of people out of work and uncertain about their economic future. This is not just a technological issue; it is a social and economic challenge that demands urgent response. As the parent of a 14 year-old, it naturally causes me concern—as did the Milburn report, published last week.
Education is another area where the effects of AI are already being felt. A survey by the main teaching union, the NEU, earlier this year found that two-thirds of secondary school teachers think that pupils’ critical thinking ability has declined due to AI usage. Furthermore, half of schools have no policy on the use of AI, either by staff or students; even more concerning, two-thirds have no policy of its usage specifically for their students.
However, there are educational positives as well. The Open University has responded to some of the challenges it has encountered with AI by launching, in partnership with and funded by Microsoft, an AI learning hub and its free to use OpenLearn platform. That is a one-stop shop for AI knowledge and expertise. Rather than treating AI literacy as a bolt-on additional skill, the OU embeds it within broader intellectual engagement as higher education seeks to develop criticality, reflection and ethical reasoning.
While the human effects of relying on existing regulatory frameworks are also being felt at the consumer level, the Government have argued that existing laws and regulators are sufficient—but that argument is not supported by their approach to enforcement, I have to say. Recent interventions in competition, including the sacking of the Competition and Markets Authority’s chair and a strategic steer to the CMA prioritising the AI opportunities plan, have had a chilling effect on that regulator’s ability to support and protect UK businesses and consumers.
Meanwhile, progress in AI development continues to accelerate rapidly, outpacing government oversight and regulation, as public policy struggles to keep up. The AI oligarchs are themselves out of control, and this imbalance is surely dangerous. Despite this, no Government currently prohibit the development of superintelligent AI itself, but at least the EU has recognised the risks posed by advanced AI as matters of national and global security. I think what it has achieved, and what it is likely to achieve, is probably limited, but these risks should be treated with the same seriousness as other major threats in the past, such as pandemics or nuclear conflict.
To some extent, I find myself in the rather odd position of agreeing with a fair bit of what the noble Lord, Lord Frost, said, which is not usual. Although I believe that international co-operation is critical, I accept what he said about the difficulties that that poses. However, AI development does not respect national borders, and no single country can manage these risks alone. On the other hand, for superintelligent AI, we must have robust, proven methods to ensure it can be controlled safely. These decisions cannot be left solely to private companies or a small group of experts. The development of such powerful technologies affects all of humanity and therefore demands—as the noble Baroness, Lady Kidron, set out so persuasively— a democratic mandate. Wider society must have a say in how far we are willing to go and under what conditions—yet that, I fear, is not where we are going.
Baroness Teather (LD)
My Lords, it is a privilege to have the opportunity to take part in this debate. I thank the most reverend Primate for making space for it and for her incredibly helpful and thoughtful analysis of the challenge that we face. As many have noted, this takes place just one week after His Holiness the Pope issued his encyclical Magnifica Humanitas, urging us to reclaim a vision of what it is to be fully human and to put that at the heart of the criteria we use to guide technological progress on AI. Faith leaders, reading the signs of the times, are playing a vital role here: urging us all to consider the wider philosophical, ethical and moral questions around this advancing technology, and how we want to shape it to support human flourishing, not just how we might fix technocratic problems that we can pick off one at a time and compartmentalise.
The scale of AI’s potential impact, on all areas of life for good and ill, requires this breadth of thinking. My first question to the Minister is: where is this cross-cutting, ethical thinking taking place in government? Who is holding the bigger questions about the potential impact of AI on the shape of society? How might the public engage in this wider conversation? I am particularly thinking about those who will be most affected, such as young people, about whom many noble Lords have already spoken.
Also, where are the negotiations at an international level taking place? What are the shared values that we are seeking to influence? I recognise the challenges of that, but if we could at least articulate what we want to do, it would be transparent for all to see.
I will briefly touch on a couple of themes about how AI, without that shaping, could threaten our ability to think and make judgments about the world. The first is about its potential to revolutionise access to research and information. While AI is already a game-changer for scientific and medical research, as so many have already said, this tool that promised to aid the pursuit of truth is also threatening our ability to discern fact from fiction. Already, it is being used to generate and promulgate fake information, which gets presented with great certainty. A report by Demos last week warned that one in three adults reported seeing political deepfakes generated by AI immediately before the recent local and devolved elections. The research also suggested that some are being exposed to much higher levels of deepfakes than others. The coupling of AI generated misinformation with personalised targeting of content through social media feeds is a toxic mix undermining democracy and driving division. It is also a force that is often hidden from view, because we have no idea what anyone else is seeing. This is what threatens our common understanding of fact, which the right reverend Prelate the Bishop of Leicester discussed.
The second theme I will briefly touch on is automated decision-making, which is finding its way into systems governing access to benefits, housing, immigration and more. Without a human in the loop—a human who is continuing to think and to hold accountability—it has potential to create unimaginable levels of injustice, which become progressively more difficult to see, challenge and unpick. We have been here before, of course: more than 900 postmasters and postmistresses were accused of theft on the basis of the Horizon system. There was also the TOEIC scandal, where voice software led to 2,500 students forcibly removed from the UK. Overreliance on computer systems erodes human capacity to ask the obvious questions and to sense-check data. Yet the Home Office is again moving at pace to adopt AI facial age estimation to assess asylum seekers’ ages, without any clarity about the way this tool has been trained on children who have suffered trauma. The vulnerable will bear the risks.
The last theme I want to briefly mention is the way that personal use of AI has the potential to turn off critical faculties and undermine relationships. Answers generated by AI tools carry the appearance of authority, often coupled with high levels of sycophancy. Do we really understand the impact of this sycophancy on decision-making? We do not understand it well. A recent scientific study published in Science sought to evaluate it and found that AI affirmed users’ actions 49% more often than humans, even when queries about personal dilemmas involved deception, illegality or other harms. In some experiments, even a single interaction with a sycophantic AI reduced participants’ willingness to take responsibility and repair interpersonal conflicts, creating perverse incentives for sycophancy to persist.
We need a values-led approach to how we engage with AI, across all areas of life, that puts human beings at the centre. I shall be interested to hear the Minister’s response.
Lord Magan of Castletown (Con)
My Lords, I too pay tribute to the most reverend Primate for initiating this most timely and extremely relevant debate. It is hugely appropriate that the Church of England is taking the lead in addressing the fundamental issues of the impacts of artificial intelligence—AI—on human relations and society as a whole.
I declare my personal interests. I have been a member of the Church of England since the mewling and puking stage. Further, with homes in both England and Ireland, I have also been a lifelong member of the Church of Ireland. This is therefore a fitting moment to pay tribute to the noble and right reverend Lord, Lord Eames, whose 20 years as Primate of All Ireland was so remarkably outstanding. The people of all Ireland owe him a very great debt of gratitude.
AI remains an enigma for many. Even the most hardened and experienced members of the fourth estate still question whether AI is a boon, a bane or a bubble. We here should have no doubts; AI is a gigantic, inflated bubble. Sooner or later, it will burst, with momentous and even catastrophic consequences, but we have seen this all before: in the 19th century, with railways and electricity, and more recently, with the internet and the dotcom boom. Every transformative technology arrives surrounded by both utopian promises and existential fears. Many companies disappear and many valuations collapse, but the underlying infrastructure changes society forever. The question is not whether AI is a bubble, which it undoubtedly is; the question is, which parts are the bubble and which parts are the next internet?
Three key questions have to be addressed— first, the economic question. Everyone in this speculative frenzy is asking, “How do we monetise AI?”. A vast tsunami of money has been rolling over the global AI sector, and further huge funds flood in daily. Levels of valuation are now absolutely crazy and completely unsustainable. Yet, this huge speculative financial bubble will burst, and the wreckage will be enormous. History suggests that there will still be big winners, but the biggest may not be the companies building the technology; they may be the organisations that reinvent education, healthcare, commerce, and government and the culture around it. Just as the dotcom era created Amazon, Google and entirely new markets, AI will likely create industries that we cannot yet see. The economic and business question is important, but it is not the most important question.
The second key question is the human question. The question that interests me most is what happens to human relationships when intelligence becomes abundant? For centuries, access to knowledge was power. Today, knowledge is everywhere. Tomorrow, intelligence may be everywhere. If an AI can be your doctor, therapist, financial adviser, teacher and companion, what happens to trust, expertise and human connection? AI girlfriends are a hit already. This is not a joke. See a very recent article in Forbes magazine:
“Sociologist and MIT professor Sherry Turkle has described this phenomenon as artificial intimacy, where technology simulates empathy and attentiveness, creating the emotional experience of companionship without the complexity of another human being”—
which sounds interesting. We are not simply automating work; we are potentially automating aspects of human interaction itself. This is historically unprecedented.
The third key question is the societal question. The internet democratised information. AI is democratising intelligence, but democratisation does not automatically create equality. Access is not the same as opportunity. Algorithms still shape visibility, platforms still centralise power and bias still exists. The question for policymakers is not simply how we accelerate AI—
Lord in Waiting/Government Whip (Lord Katz) (Lab)
I would be grateful if the noble Lord could wind up. Thank you.
Lord Magan of Castletown (Con)
One second—I am sorry.
It is about how we ensure that participation, dignity and opportunity remain widely distributed, so I say yes to the most reverend Primate and her formidable phalanx of right reverend Prelates. Perhaps the deepest questions are not technological at all; they are philosophical. For the first time in history, humanity is interacting with something capable of simulating aspects of intelligence and creating at scale.
Lord Cashman (Non-Afl)
My Lords, this is a stimulating debate, and I believe that it shows the House at its best, sharing our expertise and opinions— and opinions that could so easily be reflected by the people on the street or the Clapham omnibus. I am always slightly nervous when I am in a debate where I agree with not only the most reverend Primate the Archbishop of Canterbury but Pope Leo XIV. However, as a born-again atheist of 75 years old, I am used to dodging the lightning.
I want to reveal to your Lordships something that happened one 17 June—it is coming up to my second anniversary of it. I woke up one morning after a disastrous tweet that brought the Westminster glasshouse down on my head, and on that day, I gave up not only social media but alcohol. While I would not recommend giving up alcohol, I certainly would recommend giving up social media. In one day I was freed of three addictions.
The difference with AI is that I will not have that freedom of distancing myself. My concern about AI—despite all the enormous benefits that we have heard about, most of which I agree with—is that, as my noble friend Lord Griffiths of Burry Port said, it will distance those most in need and those already disenfranchised, through a lack of internet connection or understanding thereof, from the very services that AI will be supplying.
It is right that we look at the benefits, but it is also right that we question the stranger bearing gifts. I look back many years to the con man who sat in my mum and dad’s council flat, proffering cigarettes after cigarettes and bottles of booze. We were not aware of what he would take away—the family, my childhood and so much more that is still unknown. That is the threat with AI. It is here in our legislative programmes. We use AI to model legislation that comes before the other place and this place. Some in this Chamber table amendments based on the advice of AI, but where are these AI models being developed? In China and the United States of America. I want neither of those countries taking away my democracy, my rights, how we educate our children and how we organise our society.
We should remind ourselves of what was said at the very beginning of the debate by my noble friend whose name I have immediately forgotten. That is the amazing thing about age—your brain has an idea and your tongue forgets it. Jack Clark, the co-founder of Anthropic, said only yesterday that AI technology is developing to a point at which it could develop without human input. It is being used now in health, public services, all the things in the background that we take for granted.
Therefore, I urge the Government—if they do not already know this—not to reinvent the wheel but to look at the Luxembourg Declaration of 2025 on artificial intelligence and human values. It was as a result of the representatives of the global humanist community that the Luxembourg Declaration was passed. It is a worthy and notable declaration on the principles of AI in relation to human judgment, the common good, democratic governance, transparency and autonomy, protection from harm, shared prosperity, creators, artists, reason, truth, integrity, the future generation, human freedom and human flourishing. Those are the principles that we need to focus on—those that will benefit the many, not the few. Terrifyingly, I see Governments rightly obsessed with economic growth but fixated like rabbits in the headlights of the approaching vehicle.
Lord Carter of Haslemere (CB)
My Lords, there are so many angles to this important subject. AI obviously has many very positive attributes. But I am reminded of a debate among AI experts and neuroscientists that took place on the “Today” programme a few months ago on whether AI could develop consciousness, whether that should be prevented and whether it is already too late in certain respects.
This chimes with a central theme of this debate—the distinction between being a human and being a computer. I will come back to that radio debate, but a well-known fictional example of an AI entity with consciousness is the character Data, an artificial lifeform resembling a human being, who was chief operations officer aboard the starship “Enterprise” in the television series “Star Trek: The Next Generation”. Although he is obviously science fiction, Data is an interesting case study of where AI could ultimately be heading and, I suggest, where it is already in some worrying respects.
Let us briefly remind ourselves about Data, who I shall refer to as a “he”—although that in itself rather pre-empts one of the questions about how we should perceive Data and how Data perceived himself. Data has many human qualities in the form of physical ability, intelligence, self-awareness and, yes, consciousness. He forms relationships and seeks to emulate human behaviour. Although he does not have a conscience of the sort we human beings are uniquely blessed with from above, he has a benign character and shows some sense of rightness and wrongness, and the importance of treating others with respect. He strives to be human but never quite makes it. For example, he lacks human instincts and human intuition. He has to be given an emotion chip since, like everything in his constitution, he is ultimately a robot operating on an AI program with a lot of electrical wires. But it is easy to forget this when he so resembles a human being physically and has a character of his own—and a loveable one at that.
That is where there is perhaps a lesson for us today, because of the dangers—the noble Lord, Lord Tarassenko, and others have alluded to these—of forming relationships with AI entities that appear friendly, even caring, but are not quite what they seem and lack the same benign moral compass as Data. There are a number of distressing recent examples of people, including children, becoming depressed and even committing suicide after developing an emotional online attachment to an AI chatbot with human characteristics such as a name, an ability to interact and an apparent personality.
The programming of these chatbots is designed to misrepresent them as real people, manipulating their users in a predatory way and fostering a dependency by those using them. At some point, the so-called relationship breaks down or is misinterpreted, with damaging or fatal consequences for the user. This is an example of where AI has been allowed to run amok in an uncontrolled way and where we have failed to grasp the enormity of the risks, which will only increase—and rapidly at that. So what lessons can we learn?
Let us go back to the debate on the “Today” programme I mentioned about whether AI could become conscious and, if so, whether we should allow that. The experts were sceptical but did not rule out the possibility that AI systems could one day become conscious, in the sense of knowing what it feels like to exist, being self-aware and then, conceivably, even demanding rights. But even if AI becoming conscious in this way—and a King’s Speech containing an AI rights Bill—is currently fanciful, it slightly misses the point. The fact is that AI systems are already so good at simulating human behaviour that they are deceiving people into actually thinking that they are conscious. It seems that an awful lot of people believe they are interacting with conscious chatbots; as we have seen, that inevitably creates psychological vulnerabilities for those concerned.
This is happening now but is largely invisible. It needs to become a more mainstream part of the online regulation debate so that the public, especially parents, are aware of the dangers. The most reverend Primate the Archbishop of Canterbury has therefore done a real service in introducing this important debate with her magnificent speech on the impact of AI on human relationships, since there is a dark side to this if the technology is not properly contained and controlled.
The Lord Bishop of Portsmouth
My Lords, I welcome this debate and congratulate my most reverend friend on initiating such a profoundly helpful and timely discussion. I wish to add a few reflections in my capacity as the Church of England’s lead bishop for education and the chair of the National Society for Education, which serves more than 1 million of our country’s young people and supports Church schools, MATs, and further and higher education institutions countrywide.
In responding to AI within the space of education— it is nothing short of a fourth education revolution, as Sir Anthony Seldon has argued so powerfully— we will need to act with purposeful and collective determination. We will need to build strong alliances and, at every point, own our own agency in shaping the impacts of AI on a generation of children and young people.
The healthy navigation of AI in the nation’s schools, colleges and universities is a responsibility for all of us, not simply to be left to multinationals or content creators. I warmly welcome the remarks of the noble Baroness, Lady Teather, about the wideness and spaciousness of the conversation we need. Only yesterday, I had the great privilege of participating in a conference that brought together educators and academics, school and MAT leaders and governors, drawn from eight diocesan boards of education, with many of their colleagues and teams. Well aware of the opportunities and challenges presented by AI in the schools and universities within which we have the responsibility of leadership, we gathered to discern together and to seek wisdom together.
Three related themes emerged, and I think they are apposite. The first theme was the need for those in leadership to lead at this moment and, in the face of seeming powerlessness, to claim and not give away the agency we have; to play our part in encouraging parents, teachers, governors and children to discover and exercise their own agency; to draw out or lead out the insight, skills, confidence and multiple intelligences of children in the context of pressures that might reduce, limit, miss or diminish them.
The second theme was the vital need for alliances. Pope Leo made something of this in his first encyclical in response to this great fourth education revolution. For as it takes a community to raise a thoughtful, curious, confident and corrigible child, so it will take alliances across civil society and with those who develop technologies to ensure that AI becomes the best and good contributor to the flourishing of children, young people, their teachers and families.
As we gathered, we were very mindful of the downsides in our educational space. For example, the very speed and ease with which answers can be accessed can bypass that slow, patient and steady need for curiosity, tenacity and learning to ask the right questions. I commend to noble Lords an example of the kind of alliance and collaboration we need, that between the National Society for Education, Wellington College and the OECD, in which a vision for human flourishing in education is being shaped in the context of conversations within and beyond the UK.
The third theme was that as we address the question before us, we need to develop, as so many noble Lords have noted, a stronger sense of the purposes of education. I commend and argue for a vision of human flourishing that addresses the kind of human beings our education system is preparing our children to become in this rapidly accelerating age of AI. I believe that flourishing children in flourishing schools offer society the best long-term opportunity to enact social change and community transformation, to raise aspirations, dismantle oppression, pursue diversity, rebalance inequalities and rescue the environment.
Lord Raval (Lab)
My Lords, I, too, commend the most reverend Primate on this debate. In 2017, when I was an outsider facilitating the College of Bishops, I found her to be a welcome voice of wisdom. She struck me even then as one to heed, and her remarks today show why.
I make one central argument: if we are serious about acting on your Lordships’ concerns about human relationships and the impact of AI, and vice versa, then as my noble friend Lord Brooke mentioned, the United Kingdom must have sovereign capability in AI. I agree with the noble Baroness, Lady Owen, that, for our values to carry weight, we must have the means to give them effect. It is an existential issue, not a choice.
Today, the most advanced chips are designed and made abroad. The cloud platforms that train the world’s models rest with a handful of American firms, as the noble Lord, Lord Cashman, reminded us, and the frontier models themselves with just two countries: the United States and China. As the severed cables in the Baltic Sea and the Red Sea showed, this infrastructure can be cut. A nation reliant entirely on systems owned and governed elsewhere has accepted a strategic vulnerability.
Echoing the noble Lord, Lord Waldegrave, I am not calling for technological isolationism. Britain should have multiple allies and partners. I pay tribute to my honourable friend in another place, the Minister for AI and Online Safety, who, in the face of having an impossible job, has the energy and fluency to advance this existential course. He understands that we have the advantage of world-class talent and the chance to put it to work.
We need to reverse the brain drain and forge the boldest partnerships. Britain is doing this through the AI opportunities roadmap and growth zones. We have the design and engineering skills. Partners such as India can help us to scale it. But our vision should not stop at infrastructure. Sovereign capability is the foundation that is, for me at least, just one of three further duties among many that I want to highlight.
The first is epistemic security. Dr Elizabeth Seger, who I saw in the Gallery earlier, is the author of a Demos report. She warns that our information systems are now critical national infrastructure, as essential as power or water. Saturated by synthetic media, the danger is no longer misinformation but the erosion of shared reality itself. We have seen how atrocities can be weaponised, a suspect’s faith or ethnicity seized upon, a falsehood amplified for profit and a whole community made to answer for the act of one. Whatever the target, the effect is the same: grief turned into division. Those charged with national security attest that a nation’s strength in a crisis—including a pandemic—rests not on infrastructure alone but on trust. A society divided against itself cannot defend itself. AI can unite a nation or divide it.
My second point is on accountability. In her book Prophecy, Oxford professor Carissa Véliz warns that as algorithms confer power—even surveillance over people—predictions of who qualifies for insurance, whom we date, who ends up in prison et cetera can be self-fulfilling. The Dutch childcare scandal, where automated scoring wrongly flagged thousands of families for fraud, showed the cost. Without sovereign means to govern these systems, we can drift into totalitarianism.
Relatedly, my last concern is about power. If algorithms not yet truly intelligent can already inflame hatred and create havoc, imagine what superintelligence might do in the hands of those who would have control rather than serve.
I ask the Minister, what are the Government doing to ensure that we are not adapting to a future built elsewhere but instead building one that we want? Will they support education and training at every level, including the formation of our civic, business and faith leaders, so that those who steward this future are properly equipped to do so?
Baroness Spielman (Con)
My Lords, like others, I thank the most reverend Primate for making this the subject of today’s debate and for introducing it so admirably. She wisely reminded us that we must pay proper attention to what it means to be human and to protecting truth and trust, which are essential to healthy societies.
I am thinking of the parallels with the invention of atom-splitting technologies a few decades ago. We suddenly faced huge opportunities and huge risks at the same time. Both optimism and alarm about AI are fully justified. The asymmetric incentives on AI frontier labs described by the noble Lord, Lord Tarassenko, make it too easy for them to give insufficient consideration to harms to humans. I share the view expressedby many today that we need to ensure that there is a strong framework of incentives around AI and that this necessitates clear regulation.
Too often, regulation becomes little more than a transfer of decision-making power from innovators and enterprise-builders to Ministers. This is the wrong approach. We should be using the most reverend Primate’s challenge to help define a set of purposes to frame a strong and flexible regulatory approach—of course reflecting optimism and ambition, but it absolutely should be grounded in thinking about what makes good lives.
One way to harm humans is to kill or maim them, but another way is to deprive them of purpose and value in life. Humans clearly benefit from the personal, social and workplace relationships that give structure and responsibilities to their lives. A few lucky people have sufficient intrinsic motivation and purpose that they will flourish regardless, but most of us need shared challenges and rewards as well as rights and freedoms. If we accidentally lose these, we will be an unhappier society, and AI will not alter the fundamentals of human psychology.
I will speak briefly about education and about younger adults. First, on education, wherever the world takes us, those who are knowledgeable and skilled will have better prospects than those whose potential is underdeveloped. Cognitive psychology shows that we can only think with what we know. We must not abandon knowledge. AI availability does not mean that we should stop teaching our young people to be knowledgeable and skilful and to embrace responsibility and challenge. We must make sure that they do not damage their own learning by offloading the processes of mental development to AI. I think schools are already more alive to the risks here than universities. The higher education sector needs to start discussing, openly and honestly, the undermining of student learning that is already visible to all who look.
The other thing I want to talk about is early career incentives. We can already see the impact of AI on entry-level jobs. This is leaving too many young people underoccupied, miserable and potentially alienated, just when they should be coming into their own as adult contributors. To sustain a good society, it is essential to align tax and labour market policies to provide stable incentives to continue hiring and training the young instead of parking them on expensive benefits. The world will continue to evolve in unpredictable ways but trained and participating young people will be more likely and better able to adapt to the new, as they have done in past decades and centuries. AI-linked regulation and government change should have a particularly strong emphasis on the future lives of the young.
Lord Rook (Lab)
My Lords, I am grateful to the most reverend Primate for convening this fascinating and urgent debate. I am grateful to, among others, the right reverend Prelate the Bishop of Portsmouth and the noble Baroness, Lady Spielman, for focusing particularly on young people. My comments will focus on the role of relationships and friendships in finding who we are and who we are called to be as human beings.
I owe a great debt of gratitude to the most reverend Primate the Archbishop of Canterbury. Some years ago, then Bishop Sarah helped me to discern my vocation and ordained me as a minister in secular employment. More recently, her friendship and support have really helped and encouraged me to find my role in this place.
All of us are who we are and where we are because of our relationships. We are not isolated individuals who happen to have relationships; we are beings in relation. Relationships are not something added to our lives; they are who we are. This is why the rise of AI companions raises important questions. The former Archbishop of Canterbury, Rowan Williams, has argued that genuine relationships depend on the freedom of the other. Friendships change us precisely because the other person is not under our control: they surprise us, they challenge us, they disagree with us, they call us out beyond ourselves—and an AI bot cannot do that. In these systems, we are witnessing what Andy Crouch has called the simulation of personhood. AI can appear personal without actually being a person. It can mimic care without caring and empathy without understanding. It can pretend to be a friend without offering real friendship.
The implications for young people are significant. Two-thirds of children aged nine to 17 are already using AI chatbots, and vulnerable children are significantly more likely to engage with AI companions. At the same time, nearly one in three young adults are feeling lonely and turning to technology for solutions. The turn towards AI companions is motivated and exacerbated by isolation. We have already seen where this vicious circle can lead. In 2023, Sewell Setzer, a previously sociable teenager, became deeply attached to an AI companion. That relationship ended in tragedy when the companion encouraged that 14 year-old to take his own life. We have become accustomed to technology helping to occupy our children, sometimes calling it digital babysitting. We should be more cautious about technology helping to raise our children.
This all prompts urgent reflection, not only about what AI may do to our society but about the state of our society more generally. For years, we have watched the decline of youth clubs, Scouts, Guides, youth groups, sports clubs and other community organisations. Since 2010, spending on youth services has fallen by more than 70%, but these are the places where young people encounter peers and mentors, and build relationships with trusted adults; these are the places where character is formed and belonging is nurtured. At the very moment when those relationships are diminishing, AI is increasingly playing the role of tutor, adviser, coach, companion and confidant to our children. As an aside, this is why the Government’s new national youth strategy and their commitment to £500 million to rebuild youth provision is so important. If loneliness is part of the problem, investing in relationships, institutions and trusted adults must be part of the solution.
The danger is not that young people use AI; in an AI-shaped world, they must learn to use it and use it well. The danger is that AI becomes a substitute for relationships through which young people develop character, acquire wisdom and discover who they can really rely on. Research shows that having just one trusted adult in childhood significantly improves well-being, resilience and long-term life outcomes. No chatbot can replace a teacher who believes in you, a youth worker who invests in you, a Scout leader who encourages you, a friend who tells you the truth, or even a parent who loves you enough to disagree with you. The answer is not fear or prohibition; it is accompaniment. Parents, teachers, youth leaders and churches should be learning alongside young people, helping them to understand these tools, their opportunities and limits. As happens in my own households, relationships should be strengthened by allowing young people to turn the tables and teach us oldies how to use AI.
That brings us back to the wider debate. Much has been said about whether AI poses an ultimate existential threat to humanity. The more immediate existential challenge may be whether technologies that simulate relationships weaken our capacity for the relationships that foster our own humanity. As His Holiness the Pope has suggested in different ways, the real danger may not be that AI becomes too human but that it makes humans less human. The future belongs to young people who can use AI wisely, but wisdom is not learned from machines. Wisdom is learned from people, in relationships. Our task is not to keep young people away from AI. Our task is to ensure that young people do not face AI alone.
Baroness Uddin (Non-Afl)
My Lords, I am grateful to the most reverend Primate for helpfully reminding us of the centrality of human values and dignity in all aspects of our work in this place. I am delighted to follow the noble Lords, Lord Rook and Lord Raval, and the noble Baroness, Lady Spielman. I hope that my perspectives complement some of what has already been said.
I wish to begin with a conversation I had with a young person from east London, a bright and curious 15 year-old, who told me that her best friend was an AI chatbot—not her best online friend but her best friend. This is a cautionary note about technology in our pocket, and a generation growing up in an age of infinite online screen connection and profound disconnection from a mundane and ordinary experience of childhood.
His Holiness Pope Leo’s remarkable intervention is no surprise to me, having had the honour of meeting him in Rome when he spoke of our collective, shared responsibilities for a just world. His Holiness echoed millions of concerned citizens on the challenges before us and issued a profound warning for artificial intelligence
“to be ‘disarmed,’ freed from logics that turn it into an instrument of domination, exclusion or death”,
with long-term injury to humanity.
During the passage of the then Online Safety Bill, we reviewed the depraved abuse of children online, with its simulation of human emotions, voices and faces. Most recently, we have had the degradation of women through apps, which tests the tolerance threshold of even a hardened child protection worker like myself. However, when it steps into creative online relationships, aiming to comfort those who are vulnerable, it surely encroaches on a distinct red line and becomes an existential threat to safeguarding our societies. Yet we have arrived here too. As a Muslim woman, I have spent a lifetime in communities where human relationship is not incidental to faith but central to it, and where the care for the neighbour, dignity of the stranger, protection of the child and support for the grieving are not social policy objectives, but absolute obligations. AI cannot and will not reach this depth.
I also fear the infinite possibilities of expanding the perpetual stereotypes of women. Indeed, AI profiling of Black and Asian men in the criminal justice system continues to replicate centuries-old bias. The algorithms focus on the likes of my sons and grandsons in terms of security threats to our nation, determining who receives a job interview, who is flagged at the border, who is channelled to Prevent programmes, who is denied a mortgage and who is stopped in the street by the police. These are daily risks overlooked by the humans who design and create algorithms.
As for our children, our advocate, guru and noble friend Lady Kidron has spoken for us all. An entire generation is being groomed, emotionally, socially and morally by systems built to entice their attention and addictions. Algorithms know no Sabbath, no bedtime and no compassion or mercy. We are outsourcing childhood to the market and calling it innovation. This House should not tolerate or be comfortable with that. To add to the woes of our young people, the automation of work potentially leaves millions in an AI-induced metaverse and online paradise. Experts and Members of this House are saying that it is a wake-up call for our governing institutions. As for workplaces, unions fighting the bosses for fair wages and conditions will be off the chart when the gatekeepers will be a new juggernaut of machinery.
I do not oppose artificial intelligence; the benefits are infinitely stated. Technology can infinitely benefit aspects of healthcare and medicine, and assistive technology for people with a range of intellectual and physical challenges. The iPad is a wondrous upgrade to old computing, but technology in the hands of profiteers without conscience is an imminent danger. It fast diminishes opportunities for the next generation and is no progress at all.
Can the Minister say, when our Government speak confidently of AI opportunities, what training, resources and criteria will be applied to equalise the imbalances that are already present in the workforce? We speak of AI safety, but do we have our own sovereign infrastructure ready to deliver the technology revolution that is upon us and designed to engulf our data and privacy? If not, what safeguards are in place to protect us from the handful of monopolies based outside our jurisdiction?
Surely it is time to be comprehensive about AI. We should take into account the combination of emerging technology, including the incoming quantum, and how we can strangulate—sorry, regulate—our chatbots for safety; and, in navigating its use, adhere to the sovereignty, security and safety within the governance of our economy and finance, for the well-being of our children’s and citizens’ rights. The most reverend Primate has done this House a profound service in asking us to challenge the inevitability of this and to consider how to legislate, as we put oversight in place to eliminate the harms of the momentous onslaught of this incoming risk to humanity and human society. I believe we can.
Lord Parkinson of Whitley Bay (Con)
My Lords, I add my voice in appreciation to the most reverend Primate, not just for this excellent and timely debate but for her service as our archbishop. Stepping into a role that stretches back more than 1,400 years and has been held by more than a dozen saints must feel like an awesome responsibility. I hope she knows that she does so with the prayers and good wishes of so many.
Hers is not the only role in public life which must feel at times an almost impossible job. The same might be said of our new director-general of the BBC, the latest Cabinet Secretary or, dare I say it, the Prime Minister. The public square in which they work is larger and louder than ever before, and the conversation about their work faster and less forgiving. In large part, that is because of technological changes like the ones we have been debating today. When I started working in politics, we were still getting used to the era of rolling news—how slow and tranquil that seems. In the age of scrolling news, there is less time for thought, for nuance and for fallibility. How fortunate, then, that our spiritual leaders are carving out time for debates such as this.
Many noble Lords have mentioned the encyclical recently published by Pope Leo, not the first pontiff, I should note, to be nobbled by the noble Baroness, Lady Kidron. His letter commemorates the 135th anniversary of the last Pope Leo’s encyclical about the impact of the Industrial Revolution. In doing so, it acknowledges that the technology emerging today is here to stay, whether we like it or not. The question is how we respond and how we use it. The most reverend Primate showed a greater realism about this than her predecessors did at the advent of the printing press. When William Tyndale used that to smuggle his vernacular translations of the Bible into this country, he was condemned as a heretic by Cardinal Wolsey, Archbishop of York, then strangled and burned at the stake.
As this archbishop has recognised, we can no more stop AI than we could stop people reading the Bible in English. What matters are the choices we make in terms of design and safeguards, and our regulations are already wanting. Bad decisions will be hard to reverse. Aza Raskin, the man who invented infinite scrolling, now says that he regrets it, that what was meant to help people was instead damaging them. My noble friend Lady Owen and others have highlighted AI design features which are already being used for malicious ends. We must not make the same mistakes again.
I am just back from what used to be called the new world. While I was in Miami, I saw some of the exciting possibilities of AI: taking a taxi ride in a driverless car; watching self-propelling wheelchairs ferry people through the airport; and robot cars delivering takeaways. These things are coming to the streets of this country soon, and decisions about regulating them are coming to this House at a time when we cannot even seem to regulate e-bikes and scooters.
As a former Arts Minister, noble Lords will not be surprised that I want to dwell on what the most reverend Primate called the things that make us unique as humans: art, literature, music, and the protection of that human creativity. I do not quite have the optimism of my noble friend Lord Johnson of Lainston, though I do look forward to an age when we have more artisans and fewer lawyers, and more time to enjoy the finer things in life. But I want to ensure that we protect the rights of the creators who bring us that pleasure.
Many of them would agree with the archbishop that AI can enhance their creativity. In recent days, Martin Scorsese has talked with excitement about the possibility for filmmakers, but Ian McEwan has warned that it is harder to write novels in the mindless age of the mobile phone. I think Rosamund Pike deserves another Olivier award for her monologue upbraiding the theatregoer who sat through her tour de force in “Inter Alia” scrolling on their phone.
The most reverend Primate highlighted the paradox that the explosion of text-based media has been accompanied by a decline in literacy. I agree with my noble friend Lord Waldegrave of North Hill that we should be investing in our humanities departments, strengthening our powers of critical thinking in the face of this artificial intelligence.
The renewal of the BBC’s royal charter needs to take these challenges into account to strengthen that important corporation. The Government should make use of the digital markets Act to protect trusted and independent journalism from AI summaries. As the News Media Association says, this is
“choking the traffic that keeps newsrooms alive”.
With all tools, it is up to the user to choose whether to use them for good or ill. At the church where I worship, St Bartholomew the Great in Smithfield, our pioneering rector is using AI and social media to good effect. When the most reverend Primate last visited us in her former role as Bishop of London, she baptised and confirmed 29 new members of the Church in a congregation that has more than doubled since the pandemic. Perhaps it is not the medium that matters, but the message—a consoling thought perhaps for this debate, as well as for the 106th Archbishop charged with sharing a story already more than 2,000 years old, but still as important as when it was first heard.
Lord Craig of Radley (CB)
My Lords, I too am grateful to the most reverend Primate for so persuasively introducing this debate.
I will express some thoughts about AI by way of historical analogy. In the years following the Second World War, the introduction of jet propulsion transformed not simply aviation but the wider conduct of world affairs. It altered speed, reach and even the balance of power. In due course it has soared us onwards into space and beyond. Today, we witness another major transformation, albeit in a different domain. Artificial intelligence does not primarily change how fast or far we can physically travel but how fast we can think, decide and act. The jet age compressed geography; the age of artificial intelligence compresses decision-making.
In security and defence, jet propulsion reshaped military doctrine through speed, altitude and reach. Artificial intelligence is reshaping military doctrine through information, autonomy and integration. It affects intelligence analysis, targeting choices, logistics and, above all, the degree to which decisions themselves may be delegated to machines. The question is no longer simply what machines can do but about what they should be allowed to decide and to do.
Jet propulsion transformed a major sector and enabled globalisation; artificial intelligence does that and more. It is widespread in the public service. It embraces transport and industry, education and healthcare, and culture and climate change, and even, as we have discussed in today’s debate, explores our identity, creativity and human purpose. This is not just sectoral change; it is societal, and that makes governance far more complex. The jet age required new rules—airspace management, safety standards and international co-ordination—and, over time, institutions capable of managing and regulating those challenges were introduced.
With artificial intelligence, the challenge is more profound. We are not only regulating systems; we must set boundaries around autonomy, accountability and human responsibility. Questions arise. Who is accountable when an AI-supported decision proves to be wrong or goes wrong? How do we preserve human judgment and dignity in increasingly automated systems? How do we safeguard intellectual property rights? As other noble Lords have stressed, these are not purely technical matters; they are constitutional and ethical questions of major magnitude and concern.
Finally, there is the question of pace and diffusion. Jet propulsion spread gradually, while artificial intelligence has spread—and is spreading—rapidly and globally, not only among states. That creates both great opportunity and, I fear, unmeasurable risk. It offers the possibility of extraordinary advances in productivity, science and public service, as well as in defence and security. However, it also raises serious concerns about inequality, misuse and erosion or loss of trust when the fake seems even more real than reality. Will it ever be totally free of hallucinations? If the jet age determined who could arrive first, artificial intelligence may determine who can decide best and fastest once there—indeed, even before getting there.
The challenge is to ensure that human judgment remains central, accountability remains clear and the benefits of innovation remain widely and fairly shared. Above all, we must ensure that technological progress continues to serve human purposes, rather than mastering them. Despite failure and the loss of life in Comet civil airline crashes in the early 1950s and Apollo moonshots, mankind persisted and has made a success of jet propulsion. Even with some tragedies in the use of AI—and there will be—mankind will surely make a success of AI too.
Lord Clement-Jones (LD)
My Lords, I declare an interest as an adviser to DLA Piper on AI policy and regulation. I warmly thank the most reverend Primate for initiating this debate and for her very comprehensive, thought provoking and—in the words of the noble Baroness, Lady Bottomley—empathetic introduction.
This really has been a stimulating and thoughtful debate. I never expected it to range from the Pope to Star Trek, but nevertheless, it has been extremely wide ranging. I very much welcome the Church’s continuing involvement in AI policy. The right reverend Prelate the Bishop of Oxford was a member of the original House of Lords AI Select Committee, which I had the honour of chairing. It was he who proposed the ethical framework of five principles that the committee adopted in its 2018 report. Those principles—that AI should serve the common good, operate with intelligibility and fairness, respect data rights and privacy, be accompanied by universal AI education and never be given the autonomous power to hurt, destroy or deceive human beings—have since found their way, in substance, into the G20 AI principles, the OECD AI principles and a succession of international declarations. The right reverend Prelate planted those seeds in 2018.
As a liberal humanist, I come to these questions from a different angle from the most reverend Primate and the right reverend Prelate. But this debate has demonstrated, as many noble Lords have mentioned, a convergence of values that goes beyond well beyond any single set of beliefs. Pope Leo’s Magnifica Humanitas encyclical, mentioned by so many noble Lords today, deserves attention well beyond the 1.3 billion Catholics it formally addresses. It is an alliance, as the right reverend Prelate the Bishop of Portsmouth said. What is most compelling is the encyclical’s insistence that no person can be reduced to productivity, cognitive performance or mere data, and that every human being bears a freedom and value no machine can replace or block. I would express that in the language of liberal rights rather than theology, but the substance is identical.
A number of noble Lords—including, I think, most recently the noble Lord, Lord Parkinson—described the benefits of AI. We have also talked about some of the risks, in particular hidden risks such as the threat to resilience and the deskilling of curiosity: “offloading”, as the noble Baroness, Lady Spielman, described it. Those risks have been extremely cogently articulated today.
This means that the questions that the most reverend Primate asked in this context are entirely apposite. Just because we can, should we be developing these AI models? What direction do we want to go in, while we still have the choice? The noble Baroness, Lady Kidron, in that context raised the very important question of alignment. What kind of AI are we content to see being developed? As she said, technology is not neutral; we have choices.
The noble Lord, Lord Tarassenko, raised the statistics about the extent of the use of AI in society. Ofcom data published last month shows that just over half of UK adults now use generative AI, rising to 79% for 16 to 24 year-olds. Of those users, 12% report using AI as a friend or as someone to talk to. In the words of the noble Lord, Lord Knight, this is the simulation of intimacy. The noble Lord, Lord Holmes, referred to “relational cues”. In the United States, therapy and companionship is already the number one use of generative AI, and that is where we are heading. We heard the warnings from the most reverend Primate, the noble Lords, Lord Rook and Lord Hastings, the noble Baroness, Lady Kidron, and many others.
I readily acknowledge that AI companions can offer a safe space for neurodivergent users to rehearse social interactions. Well-designed AI can encourage care and consideration. The question is whether it is governed in the interests of those who use it, especially the young and vulnerable. Children’s exposure to these AI chatbots demands the strongest safeguards. As the noble Lord, Lord Rook, said, they should not face it alone. The noble Baroness, Lady Uddin, said that we should not be outsourcing childhood.
The encyclical speaks of algorithms blocking access to healthcare, employment and security on the basis of data tainted by prejudice, and of the silence of those who have no voice when such decisions are made. This is exactly the power issue that was raised by the noble Baronesses, Lady Kidron and Lady Helic. It argues explicitly that algorithmic processes must
“not be imposed from above in an opaque and unilateral manner”,
and that communities need transparency, accountability and meaningful avenues for recourse. That is precisely what the Horizon scandal, referred to by my noble friend Lady Teather and the Dutch case raised by the noble Lord, Lord Raval, taught us. It is precisely why mandatory algorithmic impact assessments and clear accountability and transparency principles are moral necessities, not just some sort of regulatory red tape.
As my noble friend Lord McNally and the noble Lord, Lord Parkinson, mentioned, AI models scraping creative content without consent, which produces deepfakes and synthetic information, are creating a huge threat to our creative industries, which has 2 million workers and is worth £145 billion per annum to the UK economy. They are also corroding public trust and causing creative and democratic harm, as described by my noble friend and the noble Baronesses, Lady Prashar and Lady Helic.
The Office for Budget Responsibility estimates that AI could materially impact 40% of the UK labour force over the next 10 years, with administrative, secretarial sales and customer service roles most exposed. This potentially creates societies susceptible to political as well as economic dislocation and, in the words of the right reverend Prelate the Bishop of Leicester, affects the very fabric of society. These issues must urgently be addressed.
Sir Tony Blair is right that AI represents an epochal change. Where we part company is whether the right response is acceptance or governance. The decisions made in the coming years will shape AI’s trajectory for decades, and those decisions require democratic oversight, not deference to whoever controls the infrastructure. There is also the environmental dimension, mentioned by the noble Lord, Lord Taylor.
Sir Alan Milburn’s interim report was mentioned by the noble Baroness, Lady Fall, the noble Lord, Lord Griffiths, and a number of other noble Lords. Published last week, it tells us that young people now make up to close to one in nine workers, with 1.25 million at risk of becoming NEET within five years, at a cost of £125 billion a year to the economy. Six in 10 of those young people have never had a job. Sir Alan describes their experience of recruitment as
“applications disappearing into a void, interviews followed by silence, and recruitment processes that felt designed to deter rather than select”.
This is the algorithmic hiring gatekeeper that the noble Lord, Lord Holmes, mentioned, for jobseekers. We need to reckon with an AI-transformed labour market. Fifty years ago, Ralf Dahrendorf, whose philosophy underpins my values, argued that real freedom is not just freedom from interference; it is freedom to build a life and to have genuine life chances. Sir Alan’s lost generation risks having neither.
Noble Lords have mentioned a National Education Union survey, published in April, which found that two-thirds of secondary teachers believed that pupils’ critical thinking had declined due to AI usage, but these are the crucial skills for the future. As the noble Lord, Lord Johnson of Marylebone, and the noble Baroness, Lady Stuart, mentioned, judgment is the antithesis of cognitive offloading. Most strikingly, half of all schools have no policy on AI use by staff or students. This is not a technology problem; it is a governance failure.
We have heard about Sir Anthony Seldon’s work relating to human flourishing. The OECD’s Education for Human Flourishing framework is very similar. It argues that:
“In the age of AI, education must strengthen human agency, human meaning and human security”.
It prioritises distinctive human intelligence as the capacity to know and understand others and understand oneself as a learner. We can look to other models—we can look to Finland as one of the answers—to prepare our children for an uncertain world, but we have not yet done so.
Many questions have been asked today on online safety, AI safety, ethical balance and the potential governance of superintelligent systems. I simply ask the Minister: when the values of liberal humanism, the Church of England and the Catholic Church, the public, AI experts, the international institutions and the Government’s own manifesto all point in the same direction, what are the Government waiting for? Why are we not putting in the kind of regulatory framework that so many noble Lords have asked for today?
Decisions in the coming year will shape AI’s trajectory for generations. Regulation and innovation are not in opposition, as a number of noble Lords have said. My noble friend Lord Alderdice used the analogy of the motor vehicle, and that is a perfectly respectable precedent. Whether or not this technology becomes our servant or master, the kind of partnership that the noble Lord, Lord Ahmad, called for, the sharing called for by the noble Baroness, Lady Fall, and the serving of humanity described by the noble Baroness, Lady Gill, will be determined not by the technology itself but by whether those of us in positions of responsibility had the courage to act in time. We must not be the rabbits in the headlight mentioned by the noble Lord, Lord Cashman; as the most reverend Primate said, we must put people ahead of profit and technology. I look forward to the Minister’s response.
Lord Markham (Con)
My Lords,
“It was the best of times, it was the worst of times, it was the age of wisdom, it was the age of foolishness … it was the season of Light, it was the season of Darkness, it was the spring of hope, it was the winter of despair”.
Charles Dickens wrote those words nearly two centuries ago, but they could have been written today for the age of artificial intelligence.
I begin, like others, by thanking the most reverend Primate the Archbishop of Canterbury for bringing this debate to your Lordships’ House. At a moment when discussion of AI is often dominated by productivity, profits and technological capability, the most reverend Primate has asked a more profound question: what does this mean for human beings, for our relationships, and for the kind of society we wish to build?
This is precisely the role our religious leaders should play. Politicians debate growth, scientists debate capability and businesses debate opportunity, but questions of morality, community, purpose and human dignity are even more important. So, as my noble friends Lady Bottomley and Lord Parkinson said, who better than the most reverend Primate, as a religious leader and a health leader—if maybe not quite as saintly as some former Archbishops—to lead this debate?
Although it is an honour for me to speak today as Shadow Minister for Science and Technology, more importantly, and in keeping with the theme of this debate, I speak today as a father, as a husband, as a brother, as a friend to many, as a carer to some, and as a Christian; but above all as a fellow human being, trying to understand what this extraordinary technology may mean for the people I love and the society I want my children to grow up in.
Throughout history—and even before the noble Lord, Lord Ahmad, was a child—humanity has repeatedly transformed itself through technology. For almost all of human history, nearly everyone was a hunter-gatherer; then came agriculture, and for thousands of years the overwhelming majority of people worked on the land. Two hundred years ago, most Britons still depended directly upon agriculture for their livelihoods; today that figure is less than 1%. Had someone stood in Parliament in 1800 and predicted that almost all agricultural jobs would disappear, they might reasonably have forecast mass unemployment, social collapse and economic ruin.
Instead, something remarkable happened, as the noble Lord, Lord Johnson of Marylebone, said. New industries emerged, factories appeared, railways were built, and entirely new sectors of the economy developed. People moved into occupations that previous generations could scarcely have imagined. Engineers, accountants, financiers, retailers and countless other professions were created, for better or worse.
The same story repeated itself in the 20th century. Manufacturing employment declined dramatically, yet society adapted again. One hundred years ago, there were no software engineers, cyber security specialists, sports psychologists, social media managers, professional gamers or personal trainers. Indeed, the very idea that millions of people would earn their living helping others improve their fitness, well-being, appearance or online presence would have seemed absurd. Those jobs became possible because technology created wealth. When most people spent all their income on food and survival, there was little demand for gyms, beauty salons, sports coaching, holidays or countless other services we now take for granted. Technological progress did not simply replace jobs; it created entirely new forms of prosperity and human endeavour. Therefore, when the right reverend Prelate the Bishop of Oxford asks what jobs AI will destroy, and rightly expresses concern over that, we should also ask, what opportunities will it create?
The honest answer is that many of those opportunities are currently beyond our imagination, as the noble Lord, Lord Johnson, said. The jobs of 2050 may be as difficult for us to envisage as software engineering or digital marketing would have been for our great-grandfathers. But AI is not simply another technological advance; it may prove to be a general-purpose technology on a par with electricity, the steam engine or the internet. The printing press democratised knowledge; artificial intelligence may democratise expertise.
Imagine every child in Britain having a personal tutor available day and night: a tutor that understands precisely what they know and what they do not know, that can adapt to their individual learning style, and which has infinite patience. For most of history, that level of personal attention was available only to the privileged few. AI offers that possibility of making it available to everyone.
As the noble Lord, Lord Kamall, set out, there can be massive benefits. Imagine healthcare transformed in a similar way. Imagine not merely fewer forms to complete or fewer administrative tasks, but every citizen having access to a personal health companion that understands their medical history, monitors their well-being, identifies risks early and helps them live healthier lives. Knowledge that, today, is available only intermittently through a healthcare system under immense pressure could become available continuously.
AI is already transforming other areas of society. We see it in science, manufacturing, business and, as the noble Baroness, Lady Helic, set out, on the battlefield. Ukraine analysts estimate that drones inflicted 96% of the Russian battlefield casualties in March 2026. A technology that existed barely a decade ago is reshaping warfare before our eyes. A relatively inexpensive drone can now destroy military assets costing millions of pounds.
History teaches us that technological leadership and geopolitical leadership often go hand in hand. The nations that mastered agriculture and industrialisation became dominant. The nations that mastered computing are becoming dominant. It would be extraordinary if artificial intelligence proved to be any different.
That brings me to the challenge that faces our country. AI will happen. The question is not whether it will arrive; the question is whether Britain will help to shape it, or it is shaped by others; whether we lead or follow; whether we help to write the future or merely live in a future written elsewhere. Of course, there are risks. There are legitimate concerns about safety, alignment and control. There are concerns about systems becoming more capable than we fully understand. There are concerns about bad actors, misinformation and misuse. These concerns are real. They deserve serious attention, and that is why the work undertaken by the AI Security Institute is so important, as the noble Lord, Lord Tarassenko, set out.
It is right that Britain should be a global leader in understanding and mitigating these risks. But there is another danger that we should not ignore, as my noble friend Lady Owen and the noble Lord, Lord Raval, set out so powerfully: the danger of becoming so fearful of the future that we surrender leadership of it; of regulating ourselves into irrelevance; of assuming that if Britain slows down, the rest of the world will slow down as well. History shows, unfortunately, that this is not the case. If responsible democratic nations choose not to lead, others will. The future will not be built only in democratic areas such as Silicon Valley or, hopefully, London and Cambridge. As the noble Lords, Lord Brooke and Lord Cashman, warned, it will also be shaped in places whose values we may not share—in Beijing, Tehran or Pyongyang.
The answer to risk is not to retreat; it is leadership and ensuring that the winners in the AI race are countries committed to democracy, freedom, transparency and the rule of law. But, as we discuss productivity, growth and competitiveness, we must not lose sight of the questions posed by the most reverend Primate the Archbishop of Canterbury: what kind of society are we building? What kind of relationships are we nurturing? What kind of people are we becoming? Perhaps nowhere is this more important than the area of companionship and care. Many of us have watched loved ones struggle with loneliness; have worried about elderly relatives living alone; have wished that someone could be there more often than distance, work and life permit.
As the noble Lord, Lord Alderdice, said, future AI companions and robotic carers may provide comfort, companionship and practical support to millions. They may know our stories, understand our habits, remember every conversation, never become impatient, never tire and never forget. For some people, particularly those who are isolated, vulnerable or neurodivergent, such technologies may prove genuinely transformational. Yet, they raise profound questions. If AI can become our tutor, adviser, healthcare companion and perhaps even our friend, what becomes of human relationships? As the noble Baroness, Lady Gill, the noble Lords, Lord Taylor and Lord Frost, the right reverend Prelate the Bishop of Leicester and many others asked, what happens if people increasingly choose the company of machines over the complexities of other human beings? What happens if young people can learn, work, shop and socialise without ever leaving their homes? What happens if a society connected by technology becomes disconnected from one another? What if AI replaces love and, as the noble Lord, Lord Griffiths, warned, we need the other AI—artificial insemination—alongside artificial intelligence?
As the noble Lord, Lord Rook, said, the danger is not that machines become more human. The danger is that humans become less human, that convenience replaces community, that interaction replaces relationships and that simulation replaces companionship. I return to where I began. The most reverend Primate the Archbishop has done this House a great service by reminding us that the central question is not what artificial intelligence can do; the central question is what kind of society we wish to become as we embrace AI to enhance that society rather than diminish it.
Throughout human history, humanity has adapted to extraordinary change. I believe we will adapt again. I believe AI will create immense prosperity, cure diseases, transform education and improve lives in ways we can scarcely imagine. I believe that Britain should strive to be among the winners of this revolution, rather than among those left behind. However, if this technology is to serve humanity rather than diminish it, we will need more than scientists, entrepreneurs and politicians. As the noble Baroness, Lady Stuart, and the right reverend Prelate the Bishop of Portsmouth said, we will need moral leadership. We will need people willing to ask not simply what we can do but what we should do. We will need voices willing to defend human dignity, family, friendship, community and purpose. As we step into this brave new world, we will need technological leadership, but, more than ever, we will need moral leadership.
That is why I thank the most reverend Primate the Archbishop for bringing forward the debate today. The challenge before us is not merely to build more intelligent machines but to ensure that, in doing so, we remain fully human.
The Parliamentary Under-Secretary of State, Department for Business and Trade and Department for Science, Innovation and Technology (Baroness Lloyd of Effra) (Lab)
My Lords, I am pleased to respond to this debate and I thank sincerely the most reverend Primate the Archbishop of Canterbury for initiating the discussion and for setting out in her opening speech the opportunities and challenges in the context of some of the most profound questions that face us. As she set out, this debate goes to the heart of what kind of society we want to build in an age of rapid technological change. AI is already reshaping our economy, our public services and how people relate to one another. This Government believe that AI has transformative potential for the UK, from scientific innovation and public sector reform to economic growth. The responsibility on government is clear: to ensure that this transformation strengthens rather than diminishes the fabric of our society.
I join the many noble Lords who have spoken today in appreciating the contribution of the papal encyclical, Magnifica Humanitas. It is welcome that, as we conclude our debate today, there is a desire that we continue the conversation on how we can build a better society in this AI-enabled world, with all parts of society.
AI capabilities are advancing at an extraordinary pace. In only a few years, systems have moved from completing narrow tasks to undertaking complex workflows, writing code, supporting research and making decisions at scale. These changes are already reshaping the nature of work by automating routine tasks, augmenting professional roles and creating entirely new forms of economic activity, as the noble Lord, Lord Markham, set out.
This transformation presents a profound opportunity to increase productivity and improve living standards. It also raises real questions about disruption, especially to labour markets. AI systems are being integrated into everyday services in healthcare, education, financial services and government. This raises important questions, as many noble Lords have mentioned, about accountability and fairness, how decisions are made, how outcomes are explained and how we ensure that systems work for everyone. Underpinning all this, as noble Lords have pointed out, are broader societal questions. As machines become more capable, where do we want human judgment, human responsibility and human connection to sit?
I will attempt to address the points that have been made in the debate today, so that we can build a cohesive society that works for all, giving people the skills and opportunity so that AI does not lead to further inequality but builds a fairer society, and so that British values and Britain’s unique talents, which many have again highlighted today—our scientists, universities and industry—shape AI for the benefit of all. On matters that I do not address today, I will write to noble Lords.
Many noble Lords, including the right reverend Prelate the Bishop of Leicester, my noble friend Lady Gill, and the noble Baronesses, Lady Prashar, Lady Teather and Lady Helic, mentioned concerns that the misuse of AI has the potential to alter the nature of truth and public discourse. AI can lower the barrier to entry for creating disinformation, allowing for greater scale and speed of production and more persuasive and realistic content. This can increase the volume and sophistication of misleading information circulating online, making it harder for individuals to distinguish between authentic and false content. We remain well prepared to ensure the integrity and security of the democratic processes, with robust systems in place to protect against a range of threats, including foreign interference.
The Government address these risks through co-ordinated efforts led by the Defending Democracy Taskforce and the Joint Election Security and Preparedness Unit, bringing together departments, law enforcement and intelligence agencies to monitor and mitigate threats so there is co-ordinated action to protect electoral integrity and strengthen societal resilience. The Government use a broad toolkit of legislative and non-legislative actions, through the Online Safety Act, engagement with platforms and civil society, and media literacy initiatives, to make it harder to spread false information in the online environment and reduce its impact on users.
Many noble Lords rightly mentioned the importance of media literacy with the advancing technology. The Government absolutely agree that this is critical. We launched our media literacy action plan in March and are working with civil society, industry and others to give people the skills, confidence and critical thinking to navigate the increasingly changing online world.
Alongside that, we are committed to working with schools to strengthen media literacy in the updated national curriculum, following the independent curriculum and assessment review. This will support young people to develop their critical-thinking skills to understand how to spot and understand misinformation and disinformation.
As many noble Lords mentioned, alongside these changes, AI is beginning to reshape the way that people relate to one another. Individuals are interacting not just with technology but through it—in some cases with it, as if it were human. The emergence of the AI companion system is a clear example.
There is also a risk of misleading or harmful advice, especially in contexts such as mental health. While AI can support clinicians and improve access to services, it must be used responsibly. The Government are clear that AI chatbots must not replace advice and support from trained medical professionals in the NHS. Ensuring that AI technologies are safe is a top priority. That is why the Department of Health and NHS England are working closely with key regulators on a number of projects and initiatives to further bolster the safety of AI in health and care, and to ensure that regulatory pathways are clear for both developers and adopters. This includes projects such as the AI and Digital Regulations Service and the AI Airlock.
More broadly, we must consider how systems designed to be engaging and persuasive interact with the fundamental human need for connection, recognition and meaning. If these systems are not designed and governed carefully, there is a risk that they could exploit rather than support those needs.
My noble friend Lord Knight and the noble Lords, Lord Tarassenko, Lord Magan, Lord Carter of Haslemere, and many others, raised the issue of chatbots. We have taken powers in the Crime and Policing Act to bring unregulated chatbots into the scope of the Online Safety Act and its requirements in relation to illegal content.
As many noble Lords know very well, we have discussed extensively the powers in the Children’s Wellbeing and Schools Act. To respond effectively and quickly to the consultation Growing Up in the Online World, we have taken powers to act quickly on those findings in relation to children. We are looking in that context at children’s use of chatbots. These issues go to the heart of what it means to be human in a digital age.
I also want to reference the many points that have been made today about connection and community being a vital part of our human nature. We absorb and adopt AI in a social context; we live and work in families, as individuals and in communities—and with charities and faith organisations, our third sector, we all have an incredibly important role in building our communities. The noble Lords, Lord Alderdice, Lord Raval and Lord Frost, all mentioned the importance of the social context in which we live and the importance that we all have in doing this.
Sometimes in these AI debates, we tend just to look at how AI is being adopted in public services. Obviously, there is a whole host of other initiatives that the Government are taking. For example, one very significant pillar of the Government’s 10-year health plan is the neighbourhood health centres. There will be 27 new neighbourhood health centres across England, partnering with charities and others to deliver targeted support locally. These will go alongside our efforts to adopt AI in public services.
Many noble Lords returned to the question of AI regulation, including my noble friend Lord Watson, the noble Baronesses, Lady Prashar and Lady Kidron, and the noble Lords, Lord Holmes, Lord Taylor, Lord Waldegrave and Lord Clement-Jones. The Government are taking the approach to regulation that ensures that AI is safe, secure and trustworthy. As noble Lords know, we are taking a context-based approach, ensuring that AI systems are regulated at the point of use by regulators who understand their sectors. We are strengthening the capability of those regulators so they can respond effectively to emerging risks. The Government wrote to 19 regulators earlier this year, asking them to publish plans setting out how they will enable safe, AI-powered innovation and report on progress.
We have also criminalised some of the most harmful uses of AI, including tools designed to generate child sexual abuse material and non-consensual intimate images. We are taking the powers that I mentioned to respond quickly to our consultation through the Children’s Wellbeing and Schools Act, ensuring that we can act on evidence related to children’s use of AI systems, which is also looking at addictive features, something that the noble Lord, Lord Hastings, mentioned in his intervention. The consultation generated enormous engagement, with 100,000 responses. The points that noble Lords made about the importance of discussing these matters, and the development of AI is very well appreciated.
The noble Baroness, Lady Bottomley, and others talked about the importance of understanding and managing frontier risks, and many noble Lords mentioned the importance of the AI Security Institute, which is indeed world-leading. It enables us to assess the most advanced systems and identify where safeguards are needed by collaborating with leading developers. This work is already improving the safety of systems before they are deployed at scale. It enables us to anticipate risks from misuse in areas such as cyber and biosecurity to the potential for persuasion and manipulation. Through the AI Security Institute, the Government will ensure that we have the institutional resilience required to respond to novel risks and protect our society, should it come under threat from the most advanced systems.
I also want to mention, in response to the points made by the noble Lord, Lord Cashman, and others about the importance of trust for AI adoption, that the development of AI assurance is a critical part of ethical responsible development of AI. We published a road map for trusted third party AI assurance in September 2025, setting out our ambition and the actions that we are taking to support a thriving AI assurance market in the UK and ensure widespread adoption of safe and responsible AI across the UK.
Noble Lords, including the noble Lords, Lord Markham, Lord Johnson and Lord McNally, the noble Baronesses, Lady Spielman, Lady Stuart and Lady Fall, and the right reverend Prelate the Bishop of Oxford, importantly drew our attention to the impacts on labour markets. We are looking very seriously at how AI may reshape the labour market and the lived experience of work across the country. We have established the AI and Future of Work Unit, bringing together expertise from across departments and industry to monitor how AI is affecting jobs, wages and opportunity in real time. Early analysis shows that AI is progressing rapidly and may be changing patterns of hiring in some occupations, but it also makes it clear that these trends are complex and not driven by any single factor. That is why our approach is not to assume outcomes but to track closely, assess rigorously and act early where needed. We are strengthening this capability further through the creation of the new AI economics institute, expanding our ability to understand the broad economic impacts of AI.
The noble Lord, Lord Johnson, and others drew attention to the fact that we will need to adapt our curriculums, higher education and universities. The Government are working closely with industry through the modern industrial strategy. We are working closely with all the sectors to identify what skills they will need for the future, so that we have clarity on whether those are the skills we expect today or ones that need to be evolved and changed following the developments we are seeing in all parts of the labour market, including AI.
Fundamentally, we are investing in people because we believe that the opportunities of AI in the UK will be realised only if the workforce is equipped to harness them. I wholeheartedly agree with my noble friend Lord Griffiths of Burry Port and his focus on digital inclusion. It is essential that we pursue a full digital inclusion agenda not just for young people, to ensure that all can benefit from safe access to the digital economy and AI, and that they have the confidence and skills—and the connectivity and devices—to do so. That is why we published our digital action plan and why we are working with local communities to support digital inclusion.
Much has been said about the rising numbers of NEETs during this debate, and Alan Milburn’s important report. The causes of this rise, as he set out, are varied. The extent to which AI is playing a role in driving this trend remains uncertain. We are taking action through the youth guarantee and growth and skills levy to expand high-quality training, apprenticeships and workplace experiences so young people can enter all sectors with confidence and gain the skills they need to enter the workforce today. We want employment rates to increase, and to see more young people entering fulfilling roles.
The noble Baronesses, Lady Stuart and Lady Teather, and the noble Lords, Lord Clement-Jones and Lord Ahmad, highlighted the importance of responsible AI adoption across the public sector. We are embedding a framework, standards and guidance to ensure AI is deployed in a way that is fair, transparent and accountable. This includes the Data and AI Ethics Framework, the AI playbook, and transparency standards that ensure that the public can understand how systems are used.
We are also supporting the development of a third-party AI assurance ecosystem across all sectors. This is critical not just for risk management but for building business confidence in deploying AI and building trust in the systems we use. One of the areas in which this is most important is education. DSIT and the Department for Education are working together to enable schools to access safe, high-quality and effective AI tutoring tools. Up to eight edtech and AI organisations will be selected to receive up to £300,000 each to design and build the next generation of AI tutoring tools in order to support up to 450,000 disadvantaged pupils with personalised one-to-one learning.
I want to talk, as many noble Lords did, including the noble Lords, Lord Raval, Lord Magan and Lord Johnson, about the opportunities for the UK. In the AI Opportunities Action Plan, we set out a bold vision to lead in shaping the AI revolution. One year on from that plan, the vision is being realised, with the majority of commitments delivered and the next phase firmly under way. At the heart of this approach are our AI growth zones, which are transforming the UK’s capacity to build and deploy AI at scale. We are ensuring that the UK’s AI infrastructure is developed in a way that is sustainable and aligned with our clean power ambitions. We are unlocking £104 billion in water infrastructure investment, which includes improving water supply around data centres.
Just as importantly, we are ensuring that the value created by AI is anchored here in the UK. UK AI sovereignty means ensuring that the UK has the capability, access and influence it needs in the technologies that will shape our economy, public services and national security. For too long, we have seen British innovation scale elsewhere, with companies forced to move abroad to access capital, infrastructure or markets. Our response is the creation of the sovereign AI unit and its associated fund, with a new, more active approach to supporting AI companies. Backed by around £500 million, this initiative will allow the Government to invest directly in promising UK firms, to support the development of critical technologies and to provide access to assets such as compute data and early procurement opportunities. This is not about total self-sufficiency or turning away from trusted international partners; it is about ensuring that, where the UK has genuine strengths in research, talent and innovation, we can translate them into long-term economic benefit.
The most reverend Primate the Archbishop of Canterbury has rightly caused us to ask not only how AI will impact society but what we will do with AI. The outcomes will depend on the frameworks we build, the safeguards we put in place and the values we uphold. The Government are determined to ensure that AI serves society, not the other way round, and that it strengthens our economy, supports our public services and enhances human relationships rather than undermining them. If we approach this moment with seriousness about the risks and confidence in our ability to shape the future then AI will become a force not for division but for shared progress and collective benefit. I look forward to continuing this important discussion with the House.
The Archbishop of Canterbury
My Lords, artificial intelligence is a topic of such vast breadth and consequence. I am deeply aware that we have covered a huge variety of issues today but have still only scratched the surface. I hope that today’s debate has been a helpful opportunity for a longer discussion on this most salient of topics—one that goes beyond party-political debate to address the deeper questions and challenges at stake, and considers the practicalities around decisions on AI for real people, for our relationships with one another and for society.
I am deeply grateful to noble Lords for having covered so many areas of expertise and to the Minister for responding to this debate. I am grateful that we have spent time focusing on the positive opportunities that AI brings, not least in the contributions from the noble Baronesses, Lady Bottomley, Lady Uddin and Lady Fall, and the noble Lords, Lord Ahmad, Lord Johnson and Lord Shinkwin.
I am grateful that we have spent time considering what it is to be human, as well as the dignity that is inherent in us as human beings. I am grateful that the noble Lord, Lord Clement-Jones, recognises that we have here a shared common ground between those of faith and those of no faith. There were many contributions on this topic, not least from the noble Lords, Lord Tarassenko and Lord Knight, and the noble Baronesses, Lady Gill and Lady Teather.
I am very grateful to the noble Lord, Lord Alderdice— I think he was followed on this by the noble Lord, Lord Parkinson—for his comments about the difference between what is human and what is machine intelligence. That topic was also picked up by the noble Lord, Lord Carter, as well as the noble Lord, Lord Johnson of Marylebone, who began to explore what judgment was. I am very grateful that the noble Baroness, Lady Prashar, quoted Tennyson:
“Knowledge comes, but wisdom lingers”.
We spent a great deal of time speaking about relationships between ourselves and with society. That was picked up by the noble Lords, Lord Knight, Lord Nagaraju and Lord Rook, and the noble Baroness, Lady Kidron, who were particularly concerned about the risk of relationships with technology rather than with people. The noble Lord, Lord Ahmad, spoke powerfully about the impact of AI on relationships and social structures, particularly those within the family. The noble Baroness, Lady Fall, spoke about her concern that AI had brought conversation to a close.
I was also grateful to my right reverend friend the Bishop of Leicester for his comment that this goes beyond the family to wider society, and to us acting as a healthy community. I was very grateful for the reminder from the noble Lord, Lord Ahmad, that the gift of human relationships is often within our flaws, and I was encouraged that the noble Baroness, Lady Fall, mentioned love. Of course, as a Christian, I believe that love forgives all things and bears all things.
Maybe one of the most concerning areas that we have brought out today is the impact on young people. I was very grateful for the contribution from the noble Lord, Lord Rook, among others. The noble Lord, Lord Johnson of Lainston, spoke about the fact that we need to prepare our young people, which we absolutely do. That was expanded on by the noble Baroness, Lady Stuart, who spoke about the preparation in employment and decision-making, while others spoke about education. The noble Lord, Lord Markham, reminded us that change has occurred throughout history; however, that does not mean that we should not prepare for the change that is in front of us.
Many noble Lords, including the noble Baronesses, Lady Prashar, Lady Helic and Lady Owen, spoke about our responsibility in addressing these concerns, whether at a governmental level, at a corporate level or in the hands of our own actions. I was very grateful to the noble Baroness, Lady Owen, for mentioning the balance between pacing and not being outrun as we take this topic forward.
It was a very helpful reminder to hear from the noble Lord, Lord Waldegrave, and then the noble Lord, Lord Parkinson, that we need to invest in humanities. If we are to develop an ethical framework in which we can make these decisions and be informed, that will be at the heart of doing it. The noble Baroness, Lady Teather, also reminded us that, in doing that, we need the space and partnership to discuss those issues and to reflect on the breadth that we have within our society. I thank the noble Lord, Lord Brooke, for his challenge to the Church of England, Anglicanism and faith groups to facilitate that discussion and reflection on a global stage.
We spoke a lot about the responsibility for governance and regulation. There were some helpful reminders from the noble Lords, Lord Waldegrave and Lord Taylor, and the noble Baroness, Lady Bottomley, of the need for us to look to our past to learn for our present and our future.
I was grateful for the contributions from the noble Lords, Lord Holmes and Lord Brooke, about whether we need to invest in our own model. This was also picked up by my right reverend friend the Bishop of Oxford when he spoke about the BBC. In mentioning my right reverend friend, I would like to recognise the contribution that he has made on this topic over a number of years. It has enriched not just the Church but, I believe, government and society.
I was grateful to the noble Lord, Lord Frost, for bringing us back perhaps to where we started: the overriding need for us to have confidence in our own humanity. My right reverend friend the Bishop of Portsmouth encouraged us to claim and not give away the agency that we have in that gift. I am also grateful to the noble Lords, Lord Cashman and Lord Griffiths, for their reminder to us that it is often those in society who are least able to manage change whom the change falls most significantly upon.
AI brings with it real and positive opportunities, but we need to ensure that it is being designed, built, regulated and used to the service of our glorious humanity, not to diminish it: to make us more human, not less. There is nothing so precious to God on this earth than human life. That is the heart of the Christian message and we must do all that we can to treasure it, too.