Derbyshire Constabulary has removed the officer from frontline duty and is working with prosecutors after AI tools were allegedly used to prepare material that could appear in court.
A Derbyshire Constabulary officer has been removed from frontline duty and is under criminal investigation over allegations that AI systems were used to create evidential material, meaning evidence that could be used in court, in a number of cases, the force has confirmed, in what The Guardian reports as the first known criminal case of its kind in the United Kingdom.
The investigation includes an allegation of perverting the course of justice, a serious criminal offence in England and Wales that can apply to anyone who interferes with evidence or with the course of an investigation. The Guardian, which broke the story on Thursday after an initial Financial Times report, said the officer had been pulled from operational duties while the force works with the Crown Prosecution Service, the body that decides whether criminal charges should be brought in England and Wales. The officer is unnamed, and the presumption of innocence applies to every allegation.
The case lands on a fault line the country's own police leadership had already mapped. The National Police Chiefs' Council, which brings together senior police leaders across the UK, runs a Police AI centre that had previously told forces to stop using AI tools to prepare court statements, the Guardian reported, because the systems "may not be reliable enough". The warning came first. The Derbyshire case now tests whether the institutional guardrails built around that warning, including force discipline, prosecutorial review, and the courts themselves, can catch and correct a failure before it reaches a courtroom.
What is still unknown is substantial. Derbyshire Constabulary has not said how many cases may be affected, which AI systems were used, or what role the officer played. The force has not said whether any AI-generated material has already been used in a prosecution, withdrawn, or disclosed to defence teams. The Crown Prosecution Service has not commented publicly on its review. Each of those questions has a clear owner, and the next few weeks will show whether they answer.
The construction is delicate for a reason. Police evidence in England and Wales is built on a chain of disclosure, statement, and sign-off that assumes a human author and a verifiable process of collection. An AI tool trained on general text, used to draft or summarise a witness statement, a stop-and-search record, or a forensic note, can produce output that is fluent, plausible, and partly or wholly false. That is the concern that drove the National Police Chiefs' Council pause. A criminal probe into perverting the course of justice suggests the force's own assessment of the alleged conduct was serious enough to treat the matter as a potential abuse of office, not a procedural error.
The test now belongs to the system outside Derbyshire. The Crown Prosecution Service will decide whether the cases the officer touched need to be re-examined, dropped, or re-disclosed. The courts will decide whether material generated by an AI tool can be relied on. The National Police Chiefs' Council's existing guidance, issued before this case surfaced, will be measured against a real instance rather than a hypothetical one. The question the next few weeks will answer is whether that pre-existing warning was enough, or whether the first known criminal case of its kind in the UK is also the moment a stronger rule was overdue.