By Professor Fraser Sampson, former UK Biometrics & Surveillance Camera Commissioner
In biometrics, trust is a critical asset. If we are to get the most from the technology, being able to trust how it’s used will be as important as the technology itself.
The UK has a strong track record of biometric innovation and Facial Recognition Technology (FRT) has already proved itself in a number of settings, principally retail and policing. We also have a long tradition of rebelling against new technology and not everyone is yet convinced that the police should be allowed to use it. In Scotland, the governing party’s formal opposition to the police use of live FRT has prompted some to ask whether we now trust our supermarkets more than the police chief. It’s an important point but it’s the wrong question.
Neophobia is as old as we are, so demonisation was a predictable reaction to the arrival of FRT. Fears from innovative technology usually lead to the police being called at some point, whether it’s to tackle criminal insanity caused by trains or to confiscate pinball machines; now the police are covering demonstrations against their own use of technology. But there are several reasons why singling out the police use of AI-enabled biometrics doesn’t make sense.
First, policing is but one public service among many that are managing crime and offending – all are under enormous pressure. They work as a system, not a set of discrete parts. What happens in one part of the system – whether that’s prisons, probation or policing – will be felt somewhere else. The National Police Chiefs’ Council (NPCC) estimates the early release scheme to free up prison space across England and Wales starting this week will cost policing £480m and bring a rise in crime of up to 6%. As the government unveils a new national centre to tackle disinformation, it would be a stretch to claim that the courts, prisons and probation services are fully prepared for the impact of AI-enabled criminality, while local resilience forums have yet to grasp the significance of the small nuclear reactors and data centres heading their way. The success of all these policies will depend on partner agencies exploring and exploiting the potential contribution of AI-enabled tools systemically.
Second, the police and partners will be liable for some failures to protect citizens from crime if they could have used available technology but didn’t. Legally, the state must protect the citizen from inhuman or degrading treatment – a distinct but definite obligation to carry out proper investigation and prosecution. AI-enabled capabilities like FRT have materially changed the parameters of what “proper investigation and prosecution” look like. New kinds of harm including AI-generated pornographic images and degrading Deepfakes caused parliament to pass new laws to protect us. Putting the necessary tools of investigation beyond the reach of the police means putting perpetrators and victims beyond their reach too, undermining the effect of new legislation.
Using available biometrics also impacts on international obligations. The United Nations Convention on the Rights of the Child requires the best interests of the child to be a ‘top priority’ in all decisions and actions that affect children. AI tools are effective at tackling online harm to children and police use of FRT is protecting children from known abusers. How will stopping the police from using the technology meet that overarching obligation? Signatory states have also agreed to give each other the “greatest assistance” with criminal investigations and extradition proceedings – including obtaining necessary evidence. Increasingly this will require new biometrics and the police will have to address the expectations of international partners when considering their use of available technology.
Third, biometric technology like FRT has public support, particularly where its purpose is safeguarding children, protecting against sex offending and countering terrorism. The UK government has recognised this and put ‘transformative technology’ at the heart of its policing reforms, increasing national investment in new biometrics for the police, and promising to work more closely with retailers in particular. With the citizen freely using live FRT and sharing the images with the police, it is difficult to see on what basis we could ban the police from using it themselves.
Trust in biometric technology varies according to its use as much as its users. If we’re happy to share biometric data with the police to prevent harm and promote wellbeing, it may not matter much whether they’re using their own technology or piggy backing off someone else’s.
That will need effective governance and regulation. The police made some mistakes in their piecemeal adoption of FRT and supermarkets must learn from that, but not all are taking the same approach. One fundamental error is treating FRT as a plug-and-play solution and I listed some of the other issues here. Local authorities are also running significant risks. By mixing biometrics with CCTV they are creating a Frankenstein effect for surveillance but none of this is about trusting the cops less than the council or the corner shop. The UK’s data regulator has reminded policing that trust will come from strong data governance which means senior oversight and accountability; staff training and meticulous record keeping; systems accuracy and reducing the risk of unfairness or bias.
Policing is very different from, and much more complex than retail. The UN has signalled cautious optimism for the use of AI by the police but has cautioned that it is “not just another tool”. We won’t harness the potential and mitigate the risks of FRT by regulating the police – or supermarkets or citizens – in isolation. Tackling crime with new biometrics needs a coordinated approach where all partners are enabled, empowered and equipped for their respective roles.
And if you can’t trust your police or your supermarket with new biometric technology, the problem isn’t the technology.
About the author
Professor Fraser Sampson is the former UK Biometrics and Surveillance Camera Commissioner, Professor of Governance and National Security at CENTRIC, non-executive director at Facewatch, strategic advisor to the Alan Turing Institute’s Centre for Emerging Technology and Security (CETaS), and a longtime Biometric Update columnist. He is also the author of the Springer Nature book How AI-Driven Surveillance is Changing our World: Do Robots Have the Right to Remain Silent?, which explores the legal, ethical and societal implications of AI-driven surveillance.
Article Topics
biometrics | digital trust | facial recognition | Fraser Sampson | law enforcement