Essex police have suspended live facial recognition (LFR) systems after a University of Cambridge study revealed the technology disproportionately targeted Black individuals. The Information Commissioner’s Office (ICO) confirmed the pause, urging other police forces to address accuracy and bias risks. The study, involving 188 actors passing LFR vans in Chelmsford, found Black participants were statistically more likely to be correctly identified than others, raising fairness concerns.

The ICO’s report highlighted that while LFR systems rarely misidentified people, they showed significant racial disparities. Dr. Matt Bland, study author, noted Black individuals faced higher odds of being flagged as suspects. This contrasts with public fears about false positives, such as a South Asian man wrongly arrested for a burglary. The Home Office reported 1,300 arrests linked to LFR deployments in London between 2024 and 2025, but critics argue bias undermines trust.

The Home Secretary’s plan to expand LFR vans to 50 per police force nationwide faces scrutiny. Essex’s pilot, which used marked police vehicles to test the technology, underscores the tension between crime-solving and ethical oversight. Experts warn that unchecked AI surveillance risks exacerbating systemic inequalities. Jake Hurfurt of Big Brother Watch called for halting untested systems, emphasizing that flawed tools have no place on public streets.

The ICO’s intervention signals growing pressure to regulate biometric policing. While technical fixes, like algorithm adjustments, could mitigate bias, the study’s authors stress the need for ongoing monitoring. As LFR expands, balancing effectiveness with equity remains a critical challenge for law enforcement and regulators alike.