Rail police extend face scanning to the Tube
British Transport Police is taking its live facial recognition trial onto the London Underground, starting at Victoria Underground station. The expansion moves a railway test that began at London Bridge into the Tube environment, where passenger flows and station design differ from mainline rail. Police say the aim is to assess performance and keep use lawful and proportionate, while privacy campaigners warn the system is becoming routine in public transport.What British Transport Police is testing
British Transport Police is expanding its trial of live facial recognition to the London Underground, with Victoria Underground station named as the first Tube deployment. The cameras are then expected to rotate between Underground and Network Rail stations until the trial ends in November.The force polices railways across England, Scotland and Wales. According to the report, the trial began at London Bridge railway station in February and is intended to assess how the technology performs in a railway setting. Moving the trial to Victoria changes the operating environment: the London Underground is a dense passenger network, and the source says it is estimated to handle more than 3.7 million passenger journeys a day on weekdays.
That scale is why this is more than a routine equipment test. If live facial recognition becomes normalized in major transport stations, decisions about where cameras are placed, how watchlists are built and how alerts are reviewed will affect large numbers of people who are simply travelling.
How the scan-and-alert process is described
Live facial recognition scans faces within a camera's field of view and compares them with a police watchlist. A possible match generates an alert, which an officer must review before deciding whether any further action is warranted.The report says the railway police deployment relies on the NEC NeoFace M40 algorithm, which appears to be used across several forces. That detail matters because a face recognition deployment is not only a camera placement decision. It also depends on the matching system, the quality of the image capture, the watchlist, the review process and the rules governing what happens after an alert.
The officer review step is presented as a safeguard, but it does not answer every operational question. The supplied source does not state the watchlist criteria, alert thresholds or audit process for mistaken alerts, so those remain central issues for evaluating the trial.
Police case for the Underground expansion
Chief Superintendent Chris Casey, the officer responsible for the project, said expanding deployments into London Underground stations would help assess the technology in a different transport environment while refining how it is used across the railway network. A spokesperson later referred the report back to Casey's comments.Casey also said the force's focus remains on protecting the public, preventing crime and bringing offenders to justice, while ensuring the technology is used lawfully, proportionately and transparently. That frames the deployment as a controlled trial rather than a finished permanent rollout.
The practical question is whether the trial can produce evidence strong enough to justify wider use. Useful evidence would include how often alerts occur, how many alerts are confirmed, how many people are wrongly stopped and whether passengers receive meaningful information about scanning in the station environment.
Privacy and mistaken-match concerns
Civil liberties campaigners argue that placing live facial recognition in a major Underground station makes the technology feel routine rather than exceptional. Silkie Carlo, director of Big Brother Watch, described the expansion as "disturbing and dystopian" and warned that it would capture millions of innocent people's faces.The report also notes concern over errors, including innocent people being mistaken for criminals and detained. It says members of ethnic minorities appear to be more at risk of being mistaken for someone else by facial algorithms.
Those objections are sharper in a transit setting because Underground stations are not places people can easily avoid if they need to commute, work or connect to other travel. A person passing through Victoria may have no link to a criminal investigation, yet still pass through the camera's field of view if the system is active.
Survey findings and deletion pledge
The report cites a recent Opinium survey of 2,000 UK adults commissioned by facial recognition company Face Int. It found that 69 percent believed the public should have a say in how the technology is used, 61 percent worried that errors could get people into trouble for things they had not done, and 57 percent were concerned about how facial images were stored.British Transport Police says images of people who do not match the authorized watchlist are deleted immediately and permanently. That pledge is a key part of the force's case for proportionality, because live scanning creates public concern even when no alert is generated.
The remaining issue is governance. Immediate deletion of non-matches addresses one risk, but the public debate also covers who is placed on watchlists, how mistakes are corrected, how independent oversight works and whether trial results are published in enough detail for scrutiny.
Conclusion
The confirmed development is limited but significant: British Transport Police is moving a live facial recognition trial from mainline rail into the London Underground, beginning at Victoria and rotating cameras through stations until November. The force says the trial will help refine lawful and proportionate use in a railway environment.The unresolved questions are equally clear. Accuracy, public notice, watchlist governance, officer review, retention rules and independent scrutiny will determine whether the deployment is seen as targeted policing or routine mass surveillance. The Tube trial will therefore be watched not just as a policing experiment, but as a test of how far live biometric scanning can extend into everyday public travel.
Sources
Editorial Team - CoinBotLab
