British Police Watchdog Investigates Senior Officer Over Microsoft Copilot Use in Criminal Cases
A senior police officer in Derbyshire is facing a formal gross misconduct investigation after allegedly utilizing Microsoft’s artificial intelligence assistant to make policy determinations, evaluate risks, and justify decisions in active criminal cases. The Independent Office for Police Conduct confirmed the inquiry, noting that the investigation centers on whether the officer breached professional policing standards by feeding sensitive operational details and personal data into the commercial AI system.
The inquiry was launched in August after the Derbyshire Constabulary completed an internal review and proactively referred the matter to the watchdog on June 23. While the scope of the investigation involves serious accusations, it distinguishes itself from basic administrative tasks such as using artificial intelligence for grammar and spell checking. Investigators are examining whether the officer relied on the tool for critical investigative reasoning and policy creation between May 2025 and June 2026.
Internal policies at the Derbyshire Constabulary had permitted staff to use Microsoft Copilot beginning in May 2025 to streamline administrative duties, specifically authorizing the tool for drafting reports, organizing documents, and summarizing information. However, force guidelines explicitly prohibited inputting operational data, sensitive information, or personal data into any AI platform.
Emily Barry, the Independent Office for Police Conduct regional director, emphasized the gravity of the allegations. Investigators are tasked with determining the full extent of the tool’s deployment during the fourteen-month window. The targeted officer has been formally served with a notice of gross misconduct investigation, though no final findings have been reached and the inquiry remains ongoing.
The case underscores broader international questions regarding the integration of generative artificial intelligence into law enforcement workflows. In the United States, several police departments have adopted tools like Axon’s Draft One to generate report drafts from body-worn camera audio, prompting ongoing debates regarding automation reliance and potential misinformation. Similar technology errors have previously surfaced, including an incident in Utah where an AI-generated report mistakenly stated an officer had transformed into a frog due to background audio misinterpretations, though staff caught the error before formal submission.
Data security concerns surrounding public sector deployments also mirror private sector incidents, such as a 2023 case where corporate employees inadvertently leaked confidential source code and meeting notes to generative platforms. As investigative agencies across jurisdictions evaluate productivity gains against operational risks, the Derbyshire inquiry is expected to serve as a significant precedent for regulatory oversight and the boundaries of artificial intelligence in British policing.
