A UK Police Officer Used AI to Fake Evidence. Now They Are Investigating Him.

What Happened at Derbyshire

On June 13, 2026, Derbyshire Constabulary published a statement confirming the criminal investigation. The force said it had launched an inquiry into "an allegation of perverting the course of justice after the alleged use of AI systems by an officer to create evidential material in a number of cases" ^[1].

The officer has been removed from frontline duties pending the outcome of the investigation. No arrests have been made. The force is working with the Crown Prosecution Service (CPS) on any potentially impacted cases ^[1].

The CPS confirmed it is engaging with defence teams and the courts on affected matters, but said it would not comment further while police inquiries continue ^[1].

Derbyshire Constabulary declined to give more detail about what the evidential material consisted of. Reporting on the original Financial Times coverage of the case notes that the term "evidential material" can describe witness statements, and the force's choice of words leaves that reading open. It is also consistent with photographs, body-worn camera stills, scene-of-crime diagrams, or any other document an officer would be expected to produce in the normal course of a case ^[2].

The Same Week: The Home Office Launched PoliceAI

Here is the part the news cycle almost missed. The Derbyshire statement, picked up by Sky News and the BBC on Saturday June 13, comes in the same week the UK Home Office formally stood up a new national centre for AI in policing, branded PoliceAI ^[1].

At the PoliceAI launch on Wednesday June 10, interim director Alex Murray framed the centre's mission in unambiguous terms. "Crime and technology are evolving rapidly," Murray said. "Policing must keep pace by adopting AI responsibly to catch criminals and keep people safe" ^[1].

The collision is the political economy of the week. The UK government is in the middle of a multi-year push to deploy AI tools across British policing: facial recognition, transcription of body-worn video, automated case-file summarisation, predictive risk models, AI-assisted interview analysis. The Derbyshire case is the first live test of the integrity framework that push is supposed to rest on, and the test is failing in the most basic possible way: an officer, on the inside, allegedly using the same category of tools to fabricate the material the system is supposed to police.

The case is also the first concrete test of the College of Policing's Authorised Professional Practice (APP) on AI. The APP tells officers what they can and cannot do with AI in casework. Whether the APP is enforceable, or whether it is a paper framework that disintegrates on contact with a determined officer, is now a question the Derbyshire investigation will answer.

The Statute: Perverting the Course of Justice

The charge on the table is perverting the course of justice, a common-law offence in England and Wales carrying a maximum sentence of life imprisonment. It has been used for decades against officers who planted evidence, fabricated statements, or otherwise corrupted the integrity of proceedings. The question now is whether the existing offence is the right vehicle for AI-fabricated material, or whether a new statutory frame is needed.

Existing case law turns on the intent to pervert, and on whether the act was capable of perverting the course of justice. Both elements arguably map to an officer using AI tools to generate witness statements, photographs, or other case material. The CPS has not said which subsumption it is using in the Derbyshire investigation, and the force has not said which AI tools or systems were involved.

The Crown Prosecution Service's standing disclosure guidance already requires prosecutors to disclose the provenance of any evidence that may have been generated or assisted by AI. The Forensic Science Regulator, the UK statutory body that sets quality standards for forensic evidence, has issued general guidance on AI-generated exhibits, but the standing framework is built for analytical instruments, not generative AI tools in an officer's browser tab.

If the Derbyshire officer is charged and convicted under the existing statute, the conviction will stand. If the officer is acquitted, or if the CPS decides the existing statute does not reach AI-fabricated material, the legislative gap becomes the next fight. The Integrity units of every UK force are now reading the case file in real time.

The US-Side Comparator: Heber City and the Axon Frog

The closest US-side analogue is the Heber City, Utah case from January 2026, in which a police officer using Axon's Draft One AI report-writing software produced a case narrative that claimed the officer had turned into a frog after picking up a Disney movie ^[3]. The Heber City report was the first publicly reported case of an AI police-reporting tool putting surreal hallucinations into a sworn case file.

What is different about the Derbyshire case is intent. Heber City was a hallucination: embarrassing but not criminal. Derbyshire is an allegation of a serving officer using AI to fabricate evidential material deliberately, which is the line where hallucination stops being a vendor problem and starts being a criminal-justice problem.

The two cases also share a structural feature. Both happened because AI tools were already inside the police workflow, with policy frameworks that had not yet been stress-tested. Heber City exposed the failure mode of an AI tool asked to do a job it cannot do. The Derbyshire case, if the allegation is borne out, exposes a different failure mode: an officer using an AI tool to do a job no officer should be doing at all.

For the US reader, the pattern is the same. Axon AI police report generators, Truleo body-cam analysis, and the dozens of state and local "police AI" pilots all raise the same chain-of-custody question the Derbyshire case now lands in a UK courtroom. The structural condition is the AI tool inside the police workflow with no integrity framework that has been tested against it.

The Legal Gap the Case Will Surface

Here is the gap that should worry civil-liberties lawyers, defendants' solicitors, and the Integrity units of every UK force. The UK does not currently have a statutory offence of AI-fabricated evidence. The closest fit is perverting the course of justice, which prosecutors can use, but which was drafted for physical evidence, written statements, and oral testimony, not for AI-generated artefacts whose chain of custody the existing rules do not contemplate.

Chain of custody for AI-generated exhibits is the open problem. The Forensic Science Regulator's published standards require forensic analysts to document every step of the analytical process, the equipment used, and the version of the software. A witness statement generated in a browser tab by a commercial AI service, then pasted into a case file, has no equivalent provenance record. The Crown Prosecution Service disclosure guidance assumes a human author with a known identity. An AI tool that may have drawn on training data containing case material, prior witness statements, or police intelligence from other cases is not covered.

It is also worth noting the disclosure cascade. Defence solicitors with clients convicted in cases touched by the accused officer now have grounds to seek disclosure of the underlying AI tool, the prompts used, the outputs produced, and the full case-file trail. The first disclosure ruling in any affected case will be the earliest public test of the chain-of-custody gap.

First of Its Kind. Probably Not the Last.

Sky News broke the story on June 13. The Hacker News thread on the Sky piece crossed 316 points and 146 comments within 24 hours, and the BBC confirmed the byline the same day ^[2].

There has not previously been a publicly reported UK case of a serving police officer under criminal investigation for using AI to fabricate case evidence. The Angela Lipps wrongful-arrest case in the US in 2025, the Jacksonville Beach facial recognition case earlier in 2026, and the Heber City frog-hallucination case in January 2026 are all documented AI-policing failures, but none of them is a criminal investigation of a serving officer for AI-fabricated exhibits ^[3][4].

It is reasonable to expect the Derbyshire case to surface a backlog. Defence solicitors with clients convicted in cases touched by the accused officer now have grounds to seek disclosure. The Information Commissioner's Office (ICO) has parallel jurisdiction over police use of AI in the UK, and a regulatory investigation is plausible, particularly if the case expands to other forces. The College of Policing will be under pressure to clarify whether the APP on AI is enforceable and, if so, how. The Forensic Science Regulator will be under pressure to extend the analytical-instrument framework to AI tools used in casework by non-analyst officers.

What to Watch in the Next 7 Days

• First College of Policing or CPS statement. Both bodies are likely to issue guidance or a holding statement in the next 7 days, in part to defend their own frameworks against the implicit criticism of the case.
• Disclosure cascade in Derbyshire cases. Defence solicitors with clients convicted in cases touched by the accused officer are likely to seek disclosure. A first disclosure ruling would be the earliest test of the chain-of-custody gap.
• Information Commissioner's Office (ICO) intervention. The ICO has jurisdiction over police use of AI in the UK. A regulatory investigation is plausible, particularly if the case expands to other forces.
• PoliceAI launch follow-up. The Home Office may be forced to publicly address the timing of the Derbyshire case relative to the PoliceAI launch. The political cost of saying nothing grows with every news cycle.
• Forensic Science Regulator statement. The FSR has authority over forensic-analyst standards but no clear authority over AI tools used in casework by non-analyst officers. A statement clarifying or extending that authority is a reasonable expectation.
• Defence-bar briefings. The Criminal Bar Association and the Law Society of England and Wales will be tracking the case. A formal briefing on AI-fabricated evidence from either body would be a leading indicator that the case is being treated as a category-defining event.

The Bottom Line

A UK police officer is under criminal investigation for using AI to fabricate case evidence. The story broke the same week the UK Home Office launched its national centre for AI in policing. There is no specific UK statute covering AI-fabricated exhibits. The chain-of-custody rules do not cover AI outputs. The disclosure framework assumes a human author.

For three years the UK policing-AI conversation has been about deployment. The Derbyshire case is the first hard evidence that the integrity framework around that deployment is, at best, untested. At worst, it is not a framework at all.

It is the first UK case. It is not going to be the last.

Sources

1. BBC News: "Derbyshire Police officer accused of using AI to 'create evidence'" (June 13, 2026; primary UK news report on the criminal investigation, including the PoliceAI launch and Alex Murray quote)
2. Hacker News discussion thread: "Police officer investigated for using AI to 'create evidence' in multiple cases" (316 points, 146 comments, June 13-14, 2026; sourced from the original Sky News report, with the FT add-on clarifying that "evidential material" can describe witness statements)
3. Futurism: "Cops Forced to Explain Why AI Generated Police Report Claimed Officer Transformed Into Frog" (January 2026; the US-side Heber City AI-hallucination case)
4. CNN: "Police used AI facial recognition to arrest a Tennessee woman for crimes committed in a state she says she's never visited" (March 29, 2026; the Angela Lipps facial-recognition wrongful-arrest case)