TL;DR
Alex Murray, the UK’s national lead for AI in policing, has acknowledged that AI crime-fighting tools will contain bias but says a new £115m national centre will work to minimise the risks. His comments follow revelations of built-in bias in police facial recognition systems.
Bias Is Recognised, Not Resolved
Murray, who serves as director of threat leadership at the National Crime Agency, told The Guardian that bias in policing AI — from live facial recognition to predictive tools — is an inherent challenge that requires active management rather than denial.
“If you talk about live facial recognition or predictive policing, there will be bias,” Murray said. “You need to get in the data scientists and the data engineers to clean the data, to train the model appropriately, and then to test it.”
His admission follows a December report that found retrospective facial recognition used by police operated with inadequate safeguards. The Association of Police and Crime Commissioners responded bluntly: “System failures have been known for some time, yet these were not shared with those communities affected.”
Darryl Preston, the APCC’s forensic science lead and Cambridgeshire’s police and crime commissioner, said: “It is not acceptable for technology to be used unless and until it has been thoroughly tested to eliminate bias. That clearly was not the case in this instance.”
Arms Race With Criminals
Murray framed the push for police AI as partly defensive. Criminals are already using the technology — in one case, a convicted paedophile claimed abuse images were deepfakes, forcing police to disprove the claim.
Beyond facial recognition, Murray described AI applications spanning CCTV analysis, digital device searches, and identifying fake images used by political agitators. “What took days, weeks, sometimes months can potentially take hours,” he said.
In a Bedfordshire fraud case, AI translated Romanian-language phone data from four suspects and identified incriminating evidence within weeks, securing guilty pleas. Chief Constable Trevor Rodenhurst said officers who were once suspicious of AI are now asking when they can have it.
Looking Forward
The new £115m national AI centre will centralise assessment of private-sector AI products, replacing the current system where each of the 43 forces in England and Wales makes independent purchasing decisions. A human officer will still make final decisions on AI outputs, Murray confirmed — but the speed and scale of AI-assisted policing in the UK is set to grow considerably.