Online Safety Bill

Baroness Healy of Primrose Hill Excerpts

Committee stage
Thursday 11th May 2023 Lords Chamber

[Baroness Kidron (CB)]

My Lords, I will speak to a number of amendments in this group. I want to make the point that misinformation and disinformation was probably the issue we struggled with the most in the pre-legislative committee. We recognised the extraordinary harm it did, but also—as the noble Baroness, Lady Fox, said—that there is no one great truth. However, algorithmic spread and the drip, drip, drip of material that is not based on any search criteria or expression of an opinion but simply gives you more of the same, particularly the most shocking, moves very marginal views into the mainstream.

I am concerned that our debates over the last five days have concentrated so much on content, and that the freedom we seek does not take enough account of the way in which companies currently exercise control over the information we see. Correlations such as “Men who like barbecues are also susceptible to conspiracy theories” are then exploited to spread toxic theories that end in real-world harm or political tricks that show, for example, the Democrats as a paedophile group. Only last week I saw a series of pictures, presented as “evidence”, of President Biden caught in a compromising situation that gave truth to that lie. As Maria Ressa, the Nobel Peace Prize winner for her contribution to the freedom of expression, said in her acceptance speech:

“Tech sucked up our personal experiences and data, organized it with artificial intelligence, manipulated us with it, and created behavior at a scale that brought out the worst in humanity”.


That is the background to this set of amendments that we must take seriously.

As the noble Lord, Lord Bethell, said, Amendment 52 will ensure that platforms undertake a health misinformation risk assessment and provide a clear policy on dealing with harmful, false and misleading information. I put it to the Committee that, without this requirement, we will keep the status quo in which clicks are king, not health information.

It is a particular pleasure to support the noble Lord, Lord Moylan, on his Amendments 59 and 107. Like him, I am instinctively against taking material down. There are content-neutral ways of marking or questioning material, offering alternatives and signposting to diverse sources—not only true but diverse. These can break this toxic drip feed for long enough for people to think before they share, post and make personal decisions about the health information that they are receiving.

I am not incredibly thrilled by a committee for every occasion, but since the Bill is silent on the issue of misinformation and disinformation—which clearly will be supercharged by the rise of large language data models—it would be good to give a formal role to this advisory committee, so that it can make a meaningful and formal contribution to Ofcom as it develops not only this code of conduct but all codes of conduct.

Likewise, I am very supportive of Amendment 222, which seeks independence for the chair of the advisory body. I have seen at first hand how a combination of regulatory capture and a very litigious sector with deep pockets slows down progress and transparency. While the independence of the chair should be a given, our collective lived experience would suggest otherwise. This amendment would make that requirement clear.

Finally, and in a way most importantly, Amendment 224 would allow Ofcom to consider after the effect whether the code of conduct is necessary. This strikes a balance between adding to its current workload, which we are trying not to do, and tying one hand behind its back in the future. I would be grateful to hear from the Minister why we would not give Ofcom this option as a reasonable piece of future-proofing, given that this issue will be ever more important as AI creates layers of misinformation and disinformation at scale.

[Baroness Healy of Primrose Hill (Lab)]

My Lords, I support Amendment 52, tabled by my noble friend Lady Merron. This is an important issue which must be addressed in the Bill if we are to make real progress in making the internet a safer space, not just for children but for vulnerable adults.

We have the opportunity to learn lessons from the pandemic, where misinformation had a devastating impact, spreading rapidly online like the virus and threatening to undermine the vaccine rollout. If the Government had kept their earlier promise to include protection from harmful false health content in their indicative list of harmful content that companies would have been required to address under the now removed adult safety duties, these amendments would not be necessary.

It is naive to think that platforms will behave responsibly. Currently, they are left to their own devices in how they tackle health misinformation, without appropriate regulatory oversight. They can remove it at scale or leave it completely unchecked, as illustrated by Twitter’s decision to stop enforcing its Covid-19 misinformation policies, as other noble Lords have pointed out.

It is not a question of maintaining free speech, as some might argue. It was the most vulnerable groups who suffered from the spread of misinformation online—pregnant women and the BAME community, who had higher illness rates. Studies have shown that, proportionately, more of them died, not just because they were front-line workers but because of rumours spread in the community which resulted in vaccine hesitancy, with devastating consequences. As other noble Lords have pointed out, in 2021 the Royal College of Obstetricians and Gynaecologists found that only 42% of women who had been offered the vaccine accepted it, and in October that year one in five of the most critically ill Covid patients were unvaccinated, pregnant women. That is a heartbreaking statistic.

Unfortunately, it is not just vaccine fears that are spread on the internet. Other harmful theories can affect patients with cancer, mental health issues and sexual health issues, and, most worryingly, can affect children’s health. Rumours and misinformation play on the minds of the most vulnerable. The Government have a duty to protect people, and by accepting this amendment they would go some way to addressing this.

Platforms must undertake a health misinformation risk assessment and have a clear policy on dealing with harmful, false and misleading health information in their terms of service. They have the money and the expertise to do this, and Parliament must insist. As my noble friend Lady Merron said, I do not think that the Minister can say that the false communications offence in Clause 160 will address the problem, as it covers only a user sending a knowingly false communication with the intention of causing harm. The charity Full Fact has stated that this offence will exclude most health misinformation that it monitors online.

[Lord Clement-Jones (LD)]

My Lords, this has been a very interesting debate. I absolutely agree with what the noble Baroness, Lady Kidron, said right at the beginning of her speech. This was one of the most difficult areas that the Joint Committee had to look at. I am not saying that anything that we said was particularly original. We tried to say that this issue could be partly addressed by greater media literacy, which, no doubt, we will be talking about later today; we talked about transparency of system design, and about better enforcement of service terms and conditions. But things have moved on. Clearly, many of us think that the way that the current Bill is drafted is inadequate. However, the Government did move towards proposing a committee to review misinformation and disinformation. That is welcome, but I believe that these amendments are taking the thinking and actions a step forward.

--- Later in debate ---

[Baroness Kidron (CB)]

My Lords, over the last decade, I have been in scores of schools, run dozens of workshops and spoken to literally thousands of children and young people. A lot of what I pass off as my own wisdom in this Chamber is, indeed, their wisdom. I have a couple of points, and I speak really from the perspective of children under 18 with regard to these amendments, which I fully support.

Media literacy—or digital literacy, as it is sometimes called—is not the same as e-safety. E-safety regimes concentrate on the behaviour of users. Very often, children say that what they learn in those lessons is focused on adult anxieties about predators and bullies, and when something goes wrong, they feel that they are to blame. It puts the responsibility on children. This response, which I have heard hundreds of times, normally comes up after a workshop in which we have discussed reward loops, privacy, algorithmic bias, profiling or—my own favourite—a game which reveals what is buried in terms and conditions; for example, that a company has a right to record the sound of a device or share their data with more than a thousand other companies. When young people understand the pressures that they are under and which are designed into the system, they feel much better about themselves and rather less enamoured of the services they are using. It is my experience that they then go on to make better choices for themselves.

Secondly, we have outsourced much of digital literacy to companies such as Google and Meta. They too concentrate on user behaviour, rather than looking at their own extractive policies focused on engagement and time spent. With many schools strapped for cash and expertise, this teaching is widespread. However, when I went to a Google-run assembly, children aged nine were being taught about features available only on services for those aged over 13—and nowhere was there a mention of age limits and why they are important. It cannot be right that the companies are grooming children towards their services without taking full responsibility for literacy, if that is the literacy that children are being given in school.

Thirdly, as the Government’s own 2021 media literacy strategy set out, good media literacy is one line of defence from harm. It could make a crucial difference in people making informed and safe decisions online and engaging in a more positive online debate, at the same time as understanding that online actions have consequences offline.

However, while digital literacy and, in particular, critical thinking are fundamental to a contemporary education and should be available throughout school and far beyond, they must not be used as a way of putting responsibility on the user for the company’s design decisions. I am specifically concerned that in the risk-assessment process, digital literacy is one of the ways that a company can say it has mitigated a potential risk or harm. I should like to hear from the Minister that that is an additional responsibility and not instead of responsibility.

Finally, over all these years I have always asked at the end of the session what the young people care about the most. The second most important thing is that the system should be less addictive—it should have less addiction built into it. Again, I point the Committee in the direction of the safety-by-design amendments in the name of my noble friend Lord Russell that try to get to the crux of that. They are not very exciting amendments in this debate but they get to the heart of it. However, the thing the young people most often say is, “Could you do something to get my parents to put down their phones?” I therefore ask the Minister whether he can slip something into the Bill, and indeed ask the noble Lord, Lord Grade, whether that could emerge somewhere in the guidance. That is what young people want.

[Baroness Healy of Primrose Hill (Lab)]

My Lords, I strongly support the amendments in the name of my noble friend Lord Knight and others in this group.

We cannot entirely contain harmful, misleading and dangerous content on the internet, no matter how much we strengthen the Bill. Therefore, it is imperative that we put a new duty on category 1 and category 2A services to require them to put in place measures to promote the media literacy of users so that they can use the service safely.

I know that Ofcom takes the issue of media literacy seriously, but it is regrettable that the Government have dropped their proposal for a new media literacy duty for Ofcom. So far, I see no evidence that the platforms take media literacy seriously, so they need to be made to understand that they have corporate social responsibilities towards their clients.

Good media literacy is the first line of defence from bad information and the kind of misinformation we have discussed in earlier groups. Schools are trying to prepare their pupils to understand that the internet can peddle falsehoods as well as useful facts, but they need support, as the noble Baroness, Lady Kidron, just said. We all need to increase our media literacy, especially with the increasing use of artificial intelligence, as it can make the difference between decisions based on sound evidence and decisions based on poorly informed opinions that can harm health and well-being, social cohesion and democracy.

In 2022, Ofcom found that a third of internet users are unaware of the potential for inaccurate or biased information online, and 61% of social media users who say they are confident in judging whether online content is true or false actually lack the skills to do so, as my noble friend Lord Knight, has pointed out.

Amendment 91 would mean that platforms have to instigate measures to give users an awareness and understanding of the nature and characteristics of the content that may be on the service, its potential impact and how platforms operate. That is a sensible and practical request that is not beyond the ability of companies to provide, and it will be to everyone’s benefit.