Crime and Policing Bill

Debate between Lord Blencathra and Lord Strasburger

Committee stage
Thursday 15th January 2026 Lords Chamber

[Lord Blencathra (Con)]

My Lords, I rise to oppose Amendment 379 and support most of Amendment 471, inadequate though it is. My views may not be the same as those of my noble friends on the Front Bench, of course. We all value the right to protest, but rights are not a shield for criminality. The Government and Policing Ministers have been very clear that live facial recognition is being developed and deployed as a targeted, intelligence-led tool to identify known or wanted individuals or criminals on watch lists. It is not a blanket surveillance tool of the public. The Home Office has opened a consultation and asked for stronger statutory rules and oversight precisely to ensure proportionate lawful use.

Amendment 379 would in effect tie the hands of senior officers at the very moment when targeted identification can prevent or stop serious crime. If a protest contains people who are wanted for violent offences, sexual offences or other serious crimes, the ability to identify them quickly and safely is not an abstract technicality; it is how we protect victims and uphold the rule of law. To say that demonstrations are somehow sacrosanct and must be free from tools that help catch criminals is to place form above substance. That is not to dismiss legitimate concerns about privacy and bias. We should legislate a clear statutory framework, independent oversight and robust safeguards, and I know that the Government are consulting on exactly that path.

I will want to see strong action to correct mistakes and address suggestions that it cannot tell the difference in some ethnic groups. That has to be remedied if that allegation is true. But the right response is to legislate proportionate limits and accountability, not to pre-emptively ban a narrowly targeted operational capability at protests and thereby risk letting wanted suspects slip away. For those reasons, I urge the Committee to reject Amendment 379 and instead press the Government to bring forward the statutory code and independent oversight that the public rightly expect.

Amendment 471 is a different kettle of fish—and possibly “off” fish as well. The amendment is far too liberal and fails to protect the public from out-of-control public authorities. I will explain why. As a person relieved of ministerial duties in 1997, I found myself a rather bored Back-Bencher on the Regulation of Investigatory Powers Act 2000—the famous RIPA. The Minister at the time—I think it was Alun Michael—was waxing lyrical about how it would tackle serious crime, terrorism and paedophiles. He mentioned how it would help the police, the National Crime Agency—or whatever it was called then—MI5, MI6, HMRC and a couple of other big national government departments.

We were all in agreement that it was a jolly good thing for these agencies to have that power. Then something the Minister said prompted me to table a Question on what other public bodies could use RIPA powers, and we were shocked to discover that there were actually 32, including at that time something called the egg inspectorate of MAFF, responsible for enforcing the little lion mark on eggs. Schedule 1, listing the public authorities with phone-tapping powers, has expanded a bit since those days, and it now numbers 79. However, that is not the correct number because one of the 79 entries says “every local authority”, so we can add another 317 principal local authorities to that list. I think “every government department” covers all the agencies and arm’s-length bodies under their command, so they also have access to RIPA. In other words, a worthy proposal to let some key government agencies have power to snoop on our mobile phones to detect serious crime, terrorism or paedophilia has now become available, to some extent, to hundreds and possibly thousands of public bodies.

The relevance of this is that if we agree that facial recognition technology can be extended beyond the police, immigration, the National Crime Agency, the security services and possibly a few other big government departments that are concerned with organised crime, people trafficking and immigration, I believe our civil liberties will be at stake if local authorities and some others get to use it as well. If local authorities get the power of facial recognition, I am certain that they will abuse it. A Scottish council uses RIPA to monitor dog barking. Allerdale district council, next to me in Cumbria, used it to catch someone feeding pigeons. Of course it would be brilliant, in my opinion, to catch all those carrying out anti-social behaviour, such as riding dangerously on the pavement with their bikes, not picking up dog mess or generally causing a disturbance. But that is why I think this amendment does not go far enough.

We do not need codes of practice and safeguards—we need a complete ban on all other public authorities using it until it has been tried and tested by the police and we are satisfied that it does not cause false positives and is operationally secure. Then, if it is ever extended to other public authorities, it must be solely, as proposed new subsection (1)(a) says,

“used for the purpose of preventing, detecting, or investigating serious crimes as defined under the Serious Crime Act 2007”.

If we do not have these protections, local councils will end up checking our recycling, what library books we take out and what shops and pubs we use, and will justify it by saying it will help them deliver a better spatial strategy or design services to user patterns.

I look forward to the Liberals going back to their original roots as real liberals and bringing forward a better amendment that will protect our liberties.

[Lord Strasburger (LD)]

My Lords, I rise to support Amendment 379, to which I have added my name, and to very strongly support it. But before I do, I hope the Committee will forgive me if I digress very briefly to tidy up a matter that arose in Committee on Tuesday. I made the point that the police have the duty to facilitate protest rather than prevent it, and the noble Lord, Lord Hogan-Howe, intervened to ask me where he might find a justification for that statement. Well, I have good news. I have here the National Police Chiefs’ Council’s protest operational advice document, and on page 10, under the heading “Role of the police”, it says that authorised professional practice

“identifies two duties associated with the policing of protest. Broadly these require that the police must … not prevent, hinder or restrict peaceful assembly … in certain circumstances, take reasonable steps to protect those who want to exercise their rights peacefully. Taken together, these duties (the first a negative duty, the second a positive one) are often described as an obligation to facilitate the exercise of the freedoms of assembly and expression”.

I also have here a very handy flow chart entitled “Facilitating Peaceful Protest”, and I will make it available to the noble Lord following this debate.

To return to this group, it is now eight years since South Wales Police started deploying early versions of live facial recognition technology. When it did so, the technology was extremely inaccurate and there was absolutely no legislation in place to regulate or oversee the use of this mass surveillance technology—and that is what it is.

For those noble Lords who have not had the opportunity to experience facial recognition technology, I will give a quick overview of how it is used. It currently involves a large van full of electronics being parked in a location, such as a busy shopping street, where large numbers of ordinary people will walk past going about their daily business. On the top of the van are cameras pointing in all directions; they are scanning and recording the faces of all the passers-by. The technology tries to match them to a pre-prepared watch-list, which is a set of images of people the police want to find for some reason. Throughout the many hours of the deployment, something like 20 police officers will be standing around chatting and waiting for the system to decide, rightly or wrongly, that somebody whose face matches a person on the watch-list has just walked past. Several of the otherwise unoccupied police officers then detain the target and try to determine whether it is a true match.

Big Brother Watch, which I chair, has observed many deployments of facial recognition by the Metropolitan Police, and has seen many false matches happen. As well as false positives, the system is also susceptible to false negatives, where it fails to recognise somebody who is on the watch-list, and anyone who the police would like to speak to but was not put on the watch-list can wander by undetected. The Committee can form its own view on whether this is a productive use of scarce police time and money, but one thing is clear: this is a highly intrusive mass surveillance of thousands of citizens, almost all of whom are completely innocent and should be of no interest to the police.

The UK already has one of the highest densities of CCTV cameras in the world. Facial recognition technology will in time be added to those fixed cameras in public spaces. The police, your local authority, supermarkets or whoever will be able to keep tabs on who you are and what you are doing. This technology is far more intrusive than fingerprints or DNA. Live facial recognition can capture your face and location from a distance without you having any idea it has happened. It is as if you have a barcode on your forehead that can be read without your knowledge.

The collection and retention of fingerprints is tightly regulated by the Police and Criminal Evidence Act 1984 and the Crime and Security Act 2010. Similarly, the use of DNA is strictly regulated by the Police and Criminal Evidence Act 1984 and the Protection of Freedoms Act 2012. But what regulation is there for facial recognition, the most intrusive technology of the lot? Since the first deployment in 2017, absolutely no legislation, none at all, has been introduced to control this serious threat to our privacy. As we have already heard, the phrase “facial recognition” is not mentioned once in UK legislation.

Police forces, including the Met, have had a go at writing their own rules and marking their own homework, but that is obviously not their skill set; it is the job of legislators. The police’s homemade rules vary from force to force, and nobody is monitoring what is actually happening on the ground. For example, they assure us that all images they collect that do not match someone on the watch-list are instantly and permanently destroyed to preserve the privacy of innocent passers-by, but whether that always happens cannot be verified because there is no scrutiny, as there would be with, for example, DNA. This serious legislative vacuum is not the fault of the police; it is the fault of all the Governments since 2017, who were asleep at the wheel and did nothing to control the use of this highly intrusive technology.

You might ask: “Why does it matter to me? Why should I care if the state knows where I am and what I am doing? I am an honest, law-abiding, clean-living citizen. There is nothing in my life that I need to conceal from the police, my boss or my spouse”. You might be told by advocates of mass surveillance, “If you have nothing to hide, you have nothing to fear”. Well, that claim is first attributed to the great democrat Joseph Goebbels. The Chinese state, where much of the technology for facial recognition comes from, uses it to monitor the behaviour of its citizens. It is used not just to keep track of where they are, but to assess whether they are being good citizens in accordance with the state’s definition of what a “good citizen” is.

--- Later in debate ---

[Lord Blencathra (Con)]

My Lords, I had sought to intervene on the noble Lord, Lord Strasburger, before he sat down, but the noble Lord, Lord Hogan-Howe, beat me to it. I want to ask him a simple question but, first, I am sorry that we are on different sides of this—when we served together on the snoopers’ charter Bill, we were totally united that it was a bad Bill and we worked hand in glove to amend it. Can he tell me the substantive difference between a camera and a computer watching everyone in the crowd and picking out the wanted troublemakers and those 20 policemen he talked about looking at everybody in the crowd and picking out the wanted troublemakers from their briefing or their memory? What is the real difference between them?

[Lord Strasburger (LD)]

When I observed these deployments of facial recognition and looked at the 20 policemen standing around, it occurred to me that they would probably find a lot more of the people they were looking for if they just went round to their houses and knocked on the door, rather than working on the off-chance that they might walk past them in the high street.

[Lord Clement-Jones (LD)]

My Lords, I thank my noble friend Lady Doocey for eliciting a very useful debate, as was the intention. I particularly welcome some of the comments made by the noble Lord, Lord Hogan-Howe, but say to him that a Crime and Policing Bill might possibly be the place for discussion of the use of live facial recognition in policing. Maybe we can make some progress with the Government, we hope, responding or at least giving an indication ahead of their consultation of their approach to the legislative framework around live facial recognition. I very much hope that they will take this debate on board as part of that consultation.

As my noble friend Lady Doocey clearly stated, these amendments are necessary because live facial recognition currently operates, effectively, in a legislative void, yet the police are rolling out this technology at speed. There is no explicit Act of Parliament authorising its deployment, meaning that police forces are in effect, as my noble friend Lord Strasburger indicated, writing their own rules as they go. This technology represents a fundamental shift in the relationship between citizen and state. When LFR cameras are deployed, our public spaces become biometric checkpoints where every face is indiscriminately scanned. By treating every citizen as a suspect in a permanent digital line-up, we are abandoning the presumption of innocence. The noble Baroness, Lady Jones, made that point very well. As a result, there is a clear issue of public trust.

Amendment 379 would prohibit the use of LFR during public assemblies or processions unless a specific code of practice has been formally approved by resolution of both Houses of Parliament. This is essential to protect our freedoms of expression and assembly under Articles 10 and 11 of the ECHR. The pervasive tracking capability of LFR creates what the courts have recognised as a chilling effect, as described by my noble friend Lady Doocey and the noble Baroness, Lady Jones. Law-abiding citizens are discouraged from attending protests or expressing dissenting views for fear of permanent state monitoring. We know that police forces have already used this technology to target peaceful protesters who were not wanted for any crime. People should not have to hand over their sensitive biometric data as the price of engaging in democratic processes. Without explicit parliamentary consent and an approved code of practice, we are sleepwalking into a surveillance state that bypasses democratic oversight entirely.

Amendment 471 would establish that LFR use in public spaces must be limited to narrowly defined serious cases—such as preventing major crimes or locating missing persons—and requires prior judicial authorisation specifying the scope and purpose of each deployment. The need for this oversight was made absolutely clear by the 2020 Court of Appeal ruling in R (Bridges) v Chief Constable of South Wales Police, which found LFR use unlawful due to fundamental deficiencies in the legal framework. The court identified that far too much discretion is left to individual officers regarding who ends up on a watchlist and where cameras are placed. We must replace operational discretion with judicial scrutiny.

The Government themselves now acknowledge the inadequacy of the current framework, which they describe as a “patchwork framework” and say it is

“complicated and difficult to understand”.

Well, that is at least some progress towards the Government acknowledging the situation. They say that the current framework does not provide sufficient confidence for expanded use—hear, hear. The former Biometrics and Surveillance Camera Commissioner made clear his concerns about the College of Policing guidance, questioning whether these fundamental issues require

“more than an authorised professional practice document from the College of Policing”

and instead demand parliamentary debate. The former commissioner raised a profound question:

“Is the status of the UK citizen shifting from our jealously guarded presumption of innocence to that of ‘suspected until we have proved our identity to the satisfaction of the examining officer’?”


Such a fundamental shift in the relationship between citizen and state cannot, and should not, be determined by guidance alone.

The College of Policing’s APP on LFR, while attempting to provide operational guidance, falls short of providing the robust legal framework that this technology demands. It remains non-statutory guidance that can be revised without parliamentary scrutiny, lacks enforceable standards for deployment decisions, provides insufficient detail on bias testing and mitigation requirements, and does not establish independent oversight mechanisms with real teeth.

Most critically, the guidance permits watch-list compilation based on subjective assessments without clear statutory criteria or independent review. This leaves fundamental decisions about who gets surveilled to operational discretion rather than judicial oversight. In response to the noble Lord, Lord Blencathra, who was keen on one bit of our amendment but not the other, I say that this intelligence-led tool effectively delegates it to a senior police officer and they, in a sense, have a conflict of interest. They are the ones who make the operational decisions.