Dutourdumonde - stock.adobe.com
Met pushes ahead with major facial-recognition expansion
Metropolitan Police set to roll out live facial recognition (LFR) in the West End and Soho, but critics say police are ‘rushing ahead’ without regulation
The Metropolitan Police will deploy static live facial-recognition (LFR) cameras across the West End and Soho by the end of 2026, despite increased pressure for regulation, following a six-month “pioneering pilot” in Croydon.
Previously housed in specialist vans, static LFR cameras will now be mounted on existing street furniture such as lampposts, and monitored by officers remotely, so they can be strategically repositioned to “target emerging hotspots”.
Only activated during deployments with officers present, LFR matches are made against a “bespoke, intelligence-led watchlist”, which is created less than 24 hours in advance and deleted immediately afterwards, as per the Met’s LFR policy.
The Met said while these cameras are static, they will not be permanently fixed in any one location, with officers having the ability to reposition them to emerging hotspots if they are aware of shifts in crime patterns or tactics.
Other areas of high footfall around central London will also be targeted in the LFR expansion, with plans in place to extend its use across the city from 2027.
The continued ramp up of LFR use has been heavily criticised by multiple digital rights groups, local councillors and London Assembly members.
Responding to the Met’s announcement, Jack Coulson, head of advocacy for Big Brother Watch, said: “Legislation to regulate the police’s use of facial recognition is expected in the autumn. Yet the police are rushing ahead with artificial intelligence [AI] monitoring of the public under their own rules.”
A recent UN study also found that digital surveillance – whether real or perceived – causes a chilling effect on people’s behaviour.
Calls for regulation
At present, the Met operates based on its internal LFR policy rather than a specific legal framework, instead relying on a “patchwork” of laws, such as data protection rules and common law policing powers.
The Met’s LFR policy was found to comply with human rights law after a judicial review into its LFR use across London in April 2026. Despite this, the King’s Speech in May 2026 confirmed that a legal framework will be introduced to the Police Reform Bill – which includes the creation of an independent regulatory body and specific legal frameworks for police facial recognition to ensure “use of these technologies can be justified”.
Some 11 civil society groups – including Big Brother Watch, Justice, Liberty and the Open Rights Group – responded to the King’s Speech announcement in May 2026 by pushing for regulations in facial recognition technology, with a list of “minimum, necessary protections” to protect key the public from excessive AI surveillance.
Coulson said: “Forcing people to enter a digital police line-up in the capital’s busiest and most popular destinations is an affront to the idea that you should not have to identify yourself to the police if you have done nothing wrong. To see a play, you must now pay with your privacy.”
The ‘pioneering pilot’
The Met reported that more than 470,000 people walked past the camera during the pilot, with only one false alert, which did not result in an arrest.
LFR technology has led to more than 2,000 arrests since the start of 2024, with over 170 arrests made during the pilot trials in Croydon, where static cameras were used rather than vans.
Met commissioner Mark Rowley said that the pilot delivered a “reduction in crime, and a significant fall in violence against women and girls”, adding that “the technology supports officers to target wanted criminals and registered sex offenders”.
“We have already seen the impact in Croydon, where a six-month pilot delivered over 170 arrests, a reduction in crime, and a significant fall in violence against women and girls. All these results with only one false alert among hundreds of thousands of people,” he stated. “The technology supports officers to target wanted criminals and registered sex offenders. Crucially, it is supporting officers – not replacing them.”
Public trust
While Rowley claims that “public confidence in this is clear – around 80% of Londoners support its use”, a survey of 2,000 UK adults by facial-recognition provider Face Int found a more nuanced outlook.
“It’s far too simplistic to say that people in London are for or against facial recognition,” said Face Int CEO Tony Kounnis, as his firm’s survey revealed that 66% of Londoners said facial-recognition technology is a step towards a surveillance society, and at least half surveyed expressed worry towards wrongful identification, concerns towards racial bias and how their faces are stored.
Kounnis added: “Organisations deploying FRT need to be transparent about how the technology is used, clear about the safeguards in place and willing to recognise that people expect a consultation in how facial recognition is deployed.”
Maria Theodoulou, partner at Stokoe Partnership Solicitors, echoed that increased use of LFR without safeguards in place could undermine public confidence in both policing and the rule of law: “The push for expanded facial recognition sits uneasily with the basic legal safeguards that are meant to govern state surveillance.
“Misidentification, bias and opaque algorithms create real dangers, particularly when there are limited mechanisms to challenge or scrutinise decisions. Without enforceable oversight and consequences for non-compliance, facial recognition risks becoming another unchecked surveillance tool.”
The Home Office outlined its plans to develop the use of AI and facial recognition in UK policing in January 2026. The plan included a record £140m investment funding towards PoliceAI over the next three years, a £26m investment into the development of a national facial-recognition system and £11.6m for LFR technology.
The announcement came after a whitepaper by the government, which found that deployment of new technology varied across different police forces, leading to policing “radically under-utilising technology and data”.
In Scotland, policing bodies are at least two years away from established statutory frameworks for the use of LFR. Scottish justice secretary, Neil Gray, said that Scotland will ensure that the police’s use of the technology is “lawful, effective, proportionate and grounded in respect for human rights”.
Read more about facial recognition
- Essex Police halts live facial recognition over bias and accuracy risks: LFR deployments by Essex Police will not continue until risks associated with bias and inaccuracy have been reduced.
- How police live facial recognition subtly reconfigures suspicion: A growing body of research suggests that the use of live facial recognition is reshaping police perceptions of suspicion in ways that undermine supposed human-in-the-loop protections.
- Police facial recognition trials show little evidence of benefits: In-the-wild testing of police facial recognition systems has failed to generate clear evidence of the technology’s benefits, or to assess the full range of socio-technical impacts.
