The Metropolitan Police said on 13 May 2026 that its six-month live facial recognition pilot in Croydon led to 173 arrests, with officers making one arrest every 35 minutes across 24 separate operations. The force said the system produced just one false alert during the pilot, which ran from October 2025 to March 2026.

The results put Croydon at the centre of a test the Met says it will now keep using. The pilot used static live facial recognition cameras mounted on existing street furniture at two locations at the north and south ends of Croydon’s high street, replacing the usual van-based setup and allowing officers to run the system remotely while teams were on the ground.

Lindsey Chiswick, the Met’s lead for live facial recognition, said the results showed why the technology was powerful when used carefully, openly and in the right places. She said crime in the area was down by more than 10 per cent and that the public could see the difference. The force said the pilot also led to 37 arrests for breaches of court-imposed conditions, while those detained included people wanted for kidnap, rape and serious sexual assault.

The numbers were striking even by the Met’s own standards. More than 470,000 people passed the cameras during the six-month trial, yet the force said there was only one false alert and no one has ever been arrested because of a false alert from live facial recognition. Chiswick also said the technology was helping police find people wanted by the courts, identify serious offenders quickly and direct resources where they had the biggest impact, “all with exceptional accuracy.”

The Croydon pilot matters because it was not a one-off deployment but a sustained test of whether fixed cameras can do the job previously handled by a dedicated van. The Met said 61 per cent of the offences linked to the arrests were committed in Croydon, suggesting many of the cases were not just passing detections but were tied to crime in the borough itself. On the force’s figures, crime in the area fell by 10.5 per cent compared with the same period last year, while offences linked to violence against women and girls were down by 21 per cent.

The deployment also sharpened the debate over how the technology is used. The Met said the cameras were only activated during deployments when officers were present on the ground and that the system was monitored remotely. That approach is meant to answer concerns about misuse and mistaken identifications, but the force is also leaning on the pilot’s accuracy to justify wider use. Chiswick said the Met would continue using static cameras in Croydon as part of its regular live facial recognition deployments, which she said played a vital part in keeping London safe.

For Croydon, the question is no longer whether the pilot worked. By the Met’s account, it did. The real issue now is how quickly this model moves from test to routine police practice, and whether the promised gains in arrests and crime reduction hold up once the cameras are no longer part of a pilot.