Three million faces scanned. Over 2,100 arrests. More than 100 sex offenders taken off London's streets. And the UK's critics still want to shut it down.

The ruling that changes everything for British policing

On April 21, 2026, the UK High Court dismissed a legal challenge to the Metropolitan Police's live facial recognition system, ruling it lawful and compatible with human rights. Lord Justice Holgate and Mrs Justice Farbey found the Met's policy contains clear, interlocking constraints. This is smart policy, and I think the critics have been losing this argument for years without admitting it.

The case was brought by Shaun Thompson, an anti-knife crime youth worker who was falsely identified by cameras in Croydon, alongside Big Brother Watch director Silkie Carlo. Their argument: the technology violates privacy rights under Articles 8, 10, and 11 of the European Convention on Human Rights. The court disagreed. Thompson's experience was real and genuinely troubling. But one bad outcome does not invalidate a system that is catching rapists.

I remember watching the debate around CCTV in the 1990s. The same panic, the same Orwell references, the same predictions of a police state. Britain now has more CCTV per capita than almost anywhere on earth, and most people feel safer for it. Facial recognition is CCTV with a brain. The fear is the same. The outcome will likely be too.

The numbers that the opposition refuses to sit with

Between September 2024 and September 2025, the Metropolitan Police recorded 962 arrests for offences including rape, domestic abuse, knife crime, and robbery as a direct result of live facial recognition. Over a quarter of those arrests related to violence against women and girls.

The UK's National Physical Laboratory tested the system and found it returns the correct identity in 99% of searches. Human officers doing the same face-matching task make far more mistakes. The technology is not perfect. Nothing in policing is. But it is better than the alternative.

“The question is no longer whether we should use Live Facial Recognition — it's why we would choose not to.”

— Sir Mark Rowley, Metropolitan Police Commissioner

That is a blunt line, and I think it is the right one. Around 80% of Londoners support the use of live facial recognition to keep them safe. The public is not naive. They understand the tradeoff. The people most loudly opposed tend to be those least likely to be the victims of the crimes this technology is catching.

The bias problem is real, and the government is actually fixing it

Here is where I will give the critics their due: the bias data is genuinely bad. Of the false positives generated by LFR cameras, 80% were made on Black people. That is not a footnote. That is a serious structural problem that demands a serious structural fix.

But here is what the critics skip: the Home Office has already procured a new algorithm, independently tested by the National Physical Laboratory, that can be operated at settings with no statistically significant bias across demographic groups. Operational testing began in early 2026. The system is being fixed in real time. That is what responsible deployment looks like.

The counterpunch argument goes like this: even a fixed algorithm still enables mass surveillance, and the infrastructure itself is the threat. I do not buy that. Every LFR deployment is overt, time-bound, geographically limited, and requires senior authorisation. The watchlist is deleted after each operation. This is not China's real-time mass surveillance grid. The comparison is intellectually dishonest.

Fifty vans, a national rollout, and why the scale is justified

In January 2026, Home Secretary Shabana Mahmood announced the number of live facial recognition vans would rise from 10 to 50, with a £26 million investment in a national facial recognition system. The government also pledged £115 million over three years for a National Centre for AI in Policing. This is the most significant modernisation of British policing in a generation.

Only 13 of 43 police forces in England and Wales currently use live facial recognition. That postcode lottery is indefensible. A rape victim in Manchester deserves the same investigative tools as one in London. The national rollout is not surveillance creep. It is equity.

The Information Commissioner's Office is actively auditing police forces, publishing findings, and pushing for stronger governance. A new legal framework is being built. The oversight architecture is not absent. It is being constructed in parallel with the rollout, which is exactly how technology adoption in democratic societies should work.

Would you rather the rapist stayed free?

When was the last time you heard a critic of facial recognition explain what they would put in its place? Not a vague call for 'safeguards.' An actual alternative. More officers? The UK cannot afford them. More CCTV reviewed manually? That takes weeks. The technology exists, it works, and it is catching people who would otherwise remain free to cause harm.

The High Court's ruling is not the end of this debate. Thompson is appealing. Big Brother Watch will keep fighting. That is healthy. Judicial scrutiny of police technology is exactly what a democracy should have. But the court looked at the evidence and said: this is lawful, this is proportionate, this works. I think they got it right.

The UK is not building a surveillance state. It is building a smarter police force. Tell me those are the same thing, and I will show you someone who has never been the victim of a violent crime.