The officer has been removed from frontline duties but not yet arrested, in a week when the UK also opened PoliceAI, a new national centre to oversee responsible police use of artificial intelligence.
A Derbyshire Police officer has been removed from frontline duties and is the subject of a criminal investigation for allegedly using artificial intelligence systems to "create evidential material in a number of cases," the force has confirmed. The alleged offence is perverting the course of justice. No arrests have been made, and the investigation is in its early stages.
In a statement reported by the BBC, Derbyshire Constabulary said it had opened the inquiry after concerns were raised about an officer's use of AI. The force has not named the officer, specified the AI tools allegedly used, or said how many cases might be affected, citing the live investigation. The officer's rank and length of service are not yet public.
The Crown Prosecution Service has confirmed it is working with the force and engaging with defence teams and the courts "in appropriate cases." A spokesperson said prosecutors would review material referred by the investigation and decide on next steps in line with standard criminal-justice process. Derbyshire Constabulary said it is working closely with the CPS in relation to any potentially impacted cases.
The allegation lands in the same week the UK opened PoliceAI, a new national centre for artificial intelligence in policing. The centre was launched on Wednesday with interim director Alex Murray pledging that forces would adopt AI "responsibly" to keep pace with evolving crime and technology, according to the BBC's reporting. Its stated role is to coordinate standards, share tools, and build training across forces in England and Wales.
That timing is what makes the case more than a single-officer story. PoliceAI exists to professionalise police use of AI. A criminal investigation into an officer accused of weaponising it shows, in blunt terms, the gap between institutional promise and ground-level practice. The two facts do not prove each other: the officer is accused, not convicted, and the centre is a policy body, not the employer of the officer in question. But they sit on the same week, and the contrast is hard to miss for defence lawyers, civil liberties groups, and anyone who has watched AI evidence tools spread through law enforcement without a settled admissibility doctrine.
The legal question is whether AI-generated material can ever be evidence at all, and under what conditions. In England and Wales, evidential material must be authentic, reliable, and properly obtained. If a police officer used an AI system to draft, summarise, or generate content that was then presented as part of a case file, both the chain of custody and the duty of disclosure come under pressure. Defence teams are entitled to know how material was produced, and the Crown is obliged to disclose anything that could assist the defence. An AI system trained on undisclosed data, or one whose outputs the officer cannot fully explain, complicates both duties and may also taint cases that have already concluded.
For now, the case is in its early stages. The force has said the investigation will run its course, and the CPS has said it will engage with defence teams in appropriate cases. The number and identity of cases potentially affected are not yet public, and may emerge as the inquiry progresses. What is public, for the moment, is the allegation, the force's response, and the timing: a criminal probe into AI-assisted evidence work arriving the same week the country tries to set ground rules for police use of AI.