UK Police Officer Investigated for Using AI to Fabricate Evidence in Multiple Cases
A Derbyshire police officer is under criminal investigation for allegedly using artificial intelligence to create evidential material in multiple cases, in what is believed to be the first known instance of a UK law enforcement officer accused of fabricating evidence with generative AI.
The officer, who has not been named, has been removed from frontline duties while Derbyshire Police and the Crown Prosecution Service examine the full extent of the alleged misconduct. The investigation centers on an accusation of perverting the course of justice.
“A criminal investigation has been launched 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,” Derbyshire Police told the BBC. “The force is working closely with the Crown Prosecution Service in relation to any potentially impacted cases.”
A CPS spokesperson confirmed it was “engaging with defense teams and the courts in appropriate cases.”
A Pattern, Not an Isolated Incident
The Derbyshire investigation did not emerge in a vacuum. It follows a series of developments that reveal how quickly AI tools have penetrated police work — and how poorly the safeguards have kept up.
Just days before the Derbyshire case became public, Alex Murray, interim director of the newly established national PoliceAI centre, confirmed he had instructed several police forces in England and Wales to stop using commercially available generative AI tools to prepare court statements and criminal justice documents.
The intervention came after PoliceAI determined that any technology used in the criminal justice system must meet a standard of reliability “beyond reasonable doubt” — a threshold far higher than the accuracy expected in commercial or administrative settings.
“Police.AI is not saying don’t use AI,” the centre has stated. But it is emphasizing that AI cannot be deployed “without proper governance, testing, oversight, and accountability.”
The West Midlands Precedent
The concern is not theoretical. Last year, West Midlands Police reportedly relied on AI-generated material produced by Microsoft’s Copilot that fabricated details of a historical football match involving Maccabi Tel Aviv. The false information found its way into a police dossier supporting a proposed ban on supporters attending a match.
The incident became a high-profile example of what are commonly known as AI hallucinations — situations where generative models produce convincing but entirely false information. In a policing context, the stakes are dramatically higher than a corporate memo: a fabricated detail in a witness statement or a disclosure schedule can affect the outcome of a criminal trial.
The Evidence Integrity Problem
The Derbyshire case raises questions that go far beyond the conduct of a single officer.
If police officers are using AI to generate or summarize witness statements, who is the author of that evidence? Can a defendant cross-examine an algorithm? What obligation do forces have to disclose whether AI was used in the preparation of material presented to a court?
In the US, the Advisory Committee on Evidence Rules has been actively debating the admissibility of AI-generated and AI-manipulated evidence, with a report in May 2026 examining whether existing evidentiary frameworks can accommodate synthetic content. The UK has no equivalent formal review, leaving courts to navigate these questions on a case-by-case basis.
Legal experts have pointed out that even subtle changes introduced by AI — a rephrased sentence, an omitted qualifier, an AI-suggested detail the witness never mentioned — can undermine the reliability of evidence in ways that are extremely difficult to detect after the fact.
Parallel Investigations: The Met and Palantir
The Derbyshire case is not the only instance of AI creating turbulence in UK policing. In April 2026, the Metropolitan Police launched investigations into hundreds of its own officers after deploying an AI tool built by the US data analytics company Palantir to root out rogue personnel.
The software was deployed over the course of a week, surveilling staff members using data the force already held. It unearthed rule-breaking ranging from work-from-home violations to suspected corruption and criminal allegations including rape. As a result of the software’s analysis, three officers were arrested for offenses including abuse of authority for sexual purposes, fraud, sexual assault, and misconduct in public office.
Where the Palantir deployment represents AI used by the institution to police itself, the Derbyshire case represents the opposite: an individual officer allegedly using generative AI to manufacture the very evidence the justice system depends on.
What Perverting the Course of Justice Means
The charge under investigation — perverting the course of justice — is a serious common law offense in England and Wales, carrying a maximum sentence of life imprisonment. It requires proof that an individual performed an act that had a tendency to pervert the course of justice, and that they intended that result.
The fact that prosecutors are considering this charge suggests they believe the officer’s alleged AI use was not a procedural error or a misunderstanding of policy, but a deliberate act with the potential to corrupt legal proceedings.
This also aligns with warnings from England’s High Court. In June 2025, a senior judge warned lawyers that citing AI-generated fake case law could constitute contempt of court or even criminal prosecution for perverting the course of justice. The principle — that introducing fabricated material into legal proceedings is a serious breach — now extends from the barrister’s chambers to the police station.
The Governance Gap
The UK’s PoliceAI centre was established precisely to address these emerging risks. Its creation signals recognition at the national level that AI deployment in policing requires centralized oversight. But the centre’s interim director was reportedly still giving forces stop-orders days before the Derbyshire investigation became public — an indication that many forces had already been using the tools without formal authorization or safeguards.
The situation echoes a broader pattern observed across law enforcement globally: commercially available generative AI tools are being adopted by individual officers and units without the governance frameworks that would catch misuse before it reaches a courtroom. By the time a case like Derbyshire’s surfaces, the contaminated evidence may already have been admitted in multiple proceedings.
The VinciWorks legal analysis following PoliceAI’s intervention summed up the challenge: “AI-generated content can never be accepted at face value. Every output must be treated as a starting point for verification, not as evidence in its own right.”
For the UK criminal justice system, the Derbyshire investigation is not a one-off scandal. It is the first documented case of a problem that has likely been developing quietly — and one that will demand far more than a single investigation to solve.
Sources: BBC News, The Guardian, VinciWorks legal analysis, reporting on West Midlands Police Copilot incident, Metropolitan Police Palantir deployment, and High Court rulings on AI-generated case law.