British police are claiming measurable operational results from a live facial recognition (LFR) experiment conducted in Croydon, where artificial intelligence was combined with CCTV cameras to identify wanted individuals in public spaces. Police say the technology has contributed to a significant number of arrests during deployments in Croydon and other parts of London. The trial, run by the Metropolitan Police, involved the deployment of AI-enabled cameras in Croydon town centre. The system scans faces in real time and compares them against a police watchlist of people sought in connection with criminal investigations.
According to police statements, individuals not matching the watchlist are automatically discarded by the system. Arrests linked to the use of live facial recognition include suspects wanted for violent offences, domestic abuse, theft and knife-related crimes. In several individual operations in Croydon, police reported multiple identifications and arrests within a single deployment window.
Data released around the trial indicates that tens of thousands of faces were scanned during operations in Croydon. Police argue that the scale of scanning reflects the nature of busy urban environments and emphasise that the system is designed to operate rapidly and without storing data on people not of interest. From the police perspective, the ability to locate wanted suspects in crowded public areas without requiring officers to manually identify individuals is presented as a practical efficiency gain.
Senior police officials have described the technology as a force multiplier, allowing limited numbers of officers to cover large areas and identify suspects who might otherwise evade detection. The Croydon trial has also been cited as part of a broader move toward more permanent facial recognition infrastructure, shifting from van-mounted cameras to fixed units in selected locations.
The experiment has not been without controversy. Civil liberties groups and local campaigners have questioned the proportionality of scanning large numbers of people to generate a relatively small number of arrests, and have raised concerns about oversight and long-term regulation. However, police maintain that internal safeguards, independent reviews and existing data protection frameworks provide sufficient controls.
For law enforcement, the Croydon results are being presented as evidence that AI-enabled CCTV can deliver concrete operational outcomes when used in a targeted manner. As police forces across the UK and Europe explore the role of artificial intelligence in public safety, the Croydon trial is emerging as a reference case for how live facial recognition may be integrated into routine policing.
