The UK just launched PoliceAI, a national centre dedicated to responsible artificial intelligence in policing. The timing could not be more uncomfortable. Derbyshire Police has confirmed a criminal investigation into one of its officers for allegedly using AI systems to create evidential material across multiple cases — a potential charge of perverting the course of justice, according to Sky News. The officer has been removed from frontline duties pending the outcome of the investigation.

No arrests have been made. The Crown Prosecution Service is already working to identify which cases might be affected. This isn’t about an officer cutting corners on paperwork. Creating evidence — fabricating the material that can send someone to prison — strikes at the foundation of criminal justice itself.

A CPS spokesperson confirmed prosecutors are “engaging with defence teams and the courts in appropriate cases” that may have been affected, while declining further comment as police inquiries continue. The investigation remains in its early stages, with details still limited.

Here’s what we know so far:

• Derbyshire Police confirmed a criminal investigation into alleged use of AI to create evidential material in multiple cases
• The potential offence under investigation is perverting the course of justice
• The officer has been removed from frontline duties; no arrests made
• CPS is reviewing potentially impacted cases and liaising with defence teams and courts
• The investigation is in its early stages with limited public details

The Structural Problem Nobody Has Solved

The authentication gap in digital evidence was already widening before generative AI arrived — this case exposes just how wide it has become.

Legal experts warn there is currently no foolproof method to distinguish authentic digital content from AI-generated text, audio, video, or images, according to analysis from Esquire Solutions. Courts still rely on metadata analysis, chain-of-custody documentation, and forensic specialists. Those are the same safeguards that existed before generative AI made forgery alarmingly accessible.

It is the equivalent of installing a home security system while leaving the back door propped open.

PoliceAI’s interim director has publicly stressed that policing must “adopt AI responsibly,” according to The Times. Meanwhile, an officer allegedly weaponised that same technology to manufacture proof. The case raises an unsettling question about how many exhibits in prior proceedings were never subjected to any AI-authenticity scrutiny at all.

If substantiated, this investigation becomes a potential test case for UK AI age laws and policing regulation — one that could trigger wider reviews of affected cases, accelerate demand for forensic authentication tools, and push courts toward stricter digital evidence standards. The post-deepfake era arrived in courtrooms whether they were ready or not.