London Just Scanned 80,000 Protesters’ Faces

Scanning Protesters Is Now Official Policy

The Met Police didn’t try to hide it. The day before the protests, they announced it: “This Saturday is also the first time we will be using live facial recognition as part of a protest policing operation.” ^[1]

On May 16, facial recognition cameras were set up in the London borough of Camden, scanning the faces of an estimated 80,000 people attending two separate demonstrations. The system compares faces against a watchlist of approximately 17,000 people, primarily drawn from custody images. ^[1][5]

The Met says cameras weren’t placed within protest assembly points or on march routes themselves. But “not at the protest” and “scanning protesters on the way to the protest” is a distinction without a meaningful difference when you’re running biometric checks on everyone walking through Camden that day.

Eleven arrests were made during the operation. Two men wanted for grievous bodily harm following a hit-and-run in Birmingham were caught. One woman was arrested for allegedly refusing to remove her face mask at the Palestine march. ^[1]

Read that last one again. Arrested for covering her face at a protest where police were running facial recognition. The technology creates its own justification for enforcement.

The Met’s Case: 2,500 Arrests and “Groundbreaking” Results

The Metropolitan Police aren’t backing down from facial recognition. They’re expanding it. And they have numbers to back the argument.

Since deploying LFR technology starting in 2024, the Met says it has arrested approximately 2,500 wanted individuals. Over the twelve months to September 2025, the system scanned more than 3 million faces. Director Lindsey Chiswick called the impact “groundbreaking,” citing cases including a convicted pedophile identified while walking with a young child. ^[5]

The Met claims only 10 false alerts were detected during that period, all caught by officers before any action was taken. They cite 80% public support in quarterly surveys. ^[5]

Those numbers sound clean. Too clean. Independent audits of facial recognition systems worldwide consistently show higher error rates, particularly for women and people of color. The Met’s self-reported accuracy rate has never been independently verified at the scale they’re claiming.

The Met also describes the biometric process as “a very, very fleeting engagement of two biometric templates, and then it’s destroyed, destroyed forever.” ^[5] That’s the claim. Whether “destroyed forever” survives contact with a national security request, a policy change, or a future government with different priorities is a different question entirely.

“A Frightening Escalation”

Big Brother Watch (the UK’s most prominent surveillance watchdog) didn’t wait for the cameras to go up before responding. Jake Hurfurt, Head of Research and Investigations, issued a statement on May 14: ^[3]

“Deploying live facial recognition at protests in this country is a frightening escalation. A biometric identity check cannot become a prerequisite for free speech in this country. The use of LFR at protests will put many people off expressing their views and that is a dangerous path for Britain to go down.”

Hurfurt went further: “Police already have the powers to detain anyone using violence at a protest, but treating everybody as a potential suspect is a chilling step reminiscent of authoritarian regimes, not a democracy.” ^[3]

Senior officer Jasleen Chaggar warned of the broader trajectory: “We are at risk of becoming a nation of suspects, tracked from the moment we leave our front door.” ^[5]

Big Brother Watch has been fighting facial recognition in UK courts since 2019. They supported the Bridges v. South Wales Police case, the first legal challenge to police use of facial recognition in the UK. The Court of Appeal ruled in 2020 that South Wales Police’s use of the technology was unlawful. But the Met has pressed ahead regardless, operating under what the UK’s surveillance commissioner calls a “patchwork of laws” that provides no clear legal framework for the technology. ^[4]

No Law, No Framework, No Limits

William Webster, the UK’s Biometrics and Surveillance Camera Commissioner, has been warning about this for months. His concern isn’t abstract. It’s that police are deploying a technology with no dedicated legal framework governing its use. ^[4]

“There’s no escaping that the technologies are not foolproof,” Webster said. “They will make mistakes, and the risk is that every time a mistake is made, a police force will find themselves in a court of law.” ^[4]

Right now, police facial recognition in the UK operates under a patchwork of data protection regulation, human rights law, and common law. There is no Facial Recognition Act. No specific legislation governing when police can scan crowds, what watchlists they can use, how long data is retained, or what happens when the system gets it wrong. ^[4]

Webster has called for dedicated legislation. The government hasn’t delivered one. Instead, the Home Secretary has announced plans to expand facial recognition use across UK policing.

The legal risks Webster is flagging are real. Misidentification can lead to wrongful detention, public humiliation, and potential violations of fundamental rights including privacy, freedom of movement, and freedom of association. Every deployment without a clear legal framework is a potential lawsuit. Every lawsuit is a potential precedent. And right now, no one knows which way those precedents will fall.

The Point Is the Chill

Here’s what facial recognition at protests actually does, regardless of whether it catches any criminals: it makes people think twice about showing up.

If you know your face will be scanned, compared against a police database, and that the decision about whether you’re a “match” is made by software that’s not foolproof: do you still go to the protest? Do you still bring your kids? Do you still stand in the open where the cameras can see you?

That calculus is the chilling effect. It doesn’t require a single wrongful arrest. The mere presence of the technology changes behavior. And that’s the point civil liberties groups are making: you don’t need to arrest everyone to suppress dissent. You just need people to know they’re being watched.

The right to protest anonymously isn’t a loophole. It’s a feature of democracy. Throughout history, movements from suffrage to civil rights to labor organizing relied on the ability to show up without being individually identified by the state. Facial recognition eliminates that possibility entirely.

Coming to Every Democracy

London isn’t just London. It’s a test case. What the Met Police does, other police forces study, replicate, and scale.

China has used facial recognition at protests for years, to identify and punish dissidents. Russia deploys it against opposition demonstrators. These are the countries democratic governments point to as cautionary examples. Now London is deploying the same technology, at protests, while its own surveillance commissioner warns there’s no legal framework for it.

The United States is watching. American police departments have been adopting facial recognition steadily, with over 100 agencies now having access. San Francisco banned government use of facial recognition in 2019. Other cities followed. But the bans are fragile and the technology is spreading faster than the legislation.

If a major Western democracy deploys facial recognition at protests and faces no legal consequences, the signal to every other democracy is clear: you can do it too. The conversation shifts from “should we?” to “how soon?”

What Comes Next

• Legal challenges. Big Brother Watch has a track record of taking facial recognition to court. Expect a challenge to the protest deployment specifically, building on the Bridges v. South Wales precedent.
• Expansion. The Met has signaled this isn’t a one-off. Future protests will likely see routine facial recognition deployment. It’s now “part of the toolkit.”
• Legislation (or lack of it). Commissioner Webster’s calls for a dedicated legal framework have gone unanswered. Without legislation, police set the rules themselves.
• International ripple effects. Watch for other European police forces citing London as precedent for their own protest surveillance deployments.

The Bottom Line

On May 16, 2026, London crossed a line. The Metropolitan Police scanned the faces of 80,000 people exercising their right to protest, compared them against a police database, and called it public safety.

The technology works well enough to catch some criminals. It also works well enough to identify every person at a protest. Once that capability exists and is deployed, the question isn’t whether it will be misused. It’s when.

Sources

1. LADbible: Live Facial Recognition Cameras Used for First Time at London Protests (May 2026)
2. ITV News: Facial Recognition to Be Used in Policing Operation of Protests and FA Cup Final (May 2026)
3. Big Brother Watch: Response to Live Facial Recognition at Protests (May 2026)
4. Biometric Update: UK Watchdog Warns of Legal Risks as London Police Deploy LFR at Protest (May 2026)
5. The Star: On London’s Streets, Facial Recognition Tests Balance Between Security and Liberty (May 2026)