UK rail police face scanning pilot cost £320,000 and caught nobody

The figures come from Freedom of Information requests filed by the civil campaign group Liberty Investigates, reported by The Guardian, and cover a pilot at Network Rail stations. The Register, in its account published on 30 September, said the deployments consumed almost 100 hours of police officers' time.

BTP has not stopped. The trial now runs to November and has moved from mainline stations onto the London Underground.

The system uses NEC's NeoFace M40 technology. Cameras scan people passing through a designated area and compare their faces against a police watchlist. When the software returns a candidate, an officer reviews the alert before deciding whether to stop the person. BTP says it cannot identify people who are not on a watchlist and that biometric data for everyone else is deleted immediately. It describes deployments as intelligence-led, aimed at crime hotspots where officers believe "high harm offenders" are likely to pass through.

What the campaigners say

Jasleen Chaggar, senior legal and policy officer at Big Brother Watch, told The Register the numbers would be funny if the implications were not more serious. "It's not fair to subject innocent people to intrusive identity checks during their commute, but it's even more insulting to waste almost 100 hours of officers' time and £320,000 of public money when it produces such meagre results," she said.

Thousands of people have their highly sensitive facial data processed in public spaces as they go about their daily lives, sometimes unknowingly.

That second quote comes from Sarah Simms, senior policy officer at Privacy International. She also told The Register the result undermines the claim that live facial recognition is a targeted measure. She pointed to the absence of legislation written specifically for the technology. Specific rules, she argued, are needed to set restrictions and safeguards while BTP keeps extending its use.

There is precedent for the concern. Earlier in 2026, UK police temporarily suspended deployments after independent testing raised questions about racial bias at some operating thresholds.

So far BTP's own experiment has produced a different problem. After more than half a million faces, the only person its cameras picked out was the wrong one.

Europe's other night trains

For readers who came here for travel rather than policing, the dossier's rail material sits further from the news desk. Recent headlines, which this piece cannot treat as verified facts, point to a crowded year for overnight services: European Sleeper announcing a first Nordic route, Snälltåget adding a direct Malmö to Oslo service, and Västtrafik ordering 35 Avelia Stream Nordic trainsets.

The one checkable data point in that neighbourhood is unglamorous. TUNAI, which reads live job adverts in the UK, the US, Canada and Australia around the clock, reported on 30 September that it was tracking 621,000 live adverts that night. Of those giving a date, 34,000 claimed to have been posted more than a year ago, about 6 percent.

None of this touches the substance of the facial recognition story, which is a spending and outcomes question rather than a travel one. The £320,000 bought one false alert. Almost 100 officer hours went into the deployments. The trial now covers a larger network than when it started, and the legal framework the critics want still does not exist.

BTP's position is that the pilot was targeted and that non-watchlist faces are deleted. Its own numbers, as reported, do not show a correct match.