Microsoft Copilot Under Fresh Scrutiny After Major AI Blunder Involving the UK Police

Microsoft is already facing backlash for pushing Copilot into nearly every corner of Windows 11. More recently, reports hint about new Copilot feature coming to File Explorer. While the company and its CEO continue to argue that Copilot is genuinely useful and meant to help with everyday tasks, the AI assistant has now handed yet another reason to crictics to hate it. More importantly, the reason is hard to ignore.

As reported by The Verge, Microsoft’s Copilot was recently linked to a serious real-world mistake involving law enforcement in the UK. The issue came to light after the chief constable of West Midlands Police admitted that Copilot had hallucinated a football match that never happened, and that false information ended up inside an official intelligence report.

The report incorrectly mentioned a match between West Ham and Maccabi Tel Aviv. Just to be clear, that game doesn’t exist. However, the fabricated detail was included in police intelligence and later played a role in classifying a Europa League match involving Maccabi Tel Aviv as “high risk.”

As a result, Israeli fans were banned from attending a match against Aston Villa last year.
What makes the situation worse is that West Midlands Police had previously denied using AI in the report, initially blaming social media scraping for the error. That claim was later reversed. In a letter sent to the UK Home Affairs Committee, the chief constable confirmed that the mistake came directly from Microsoft Copilot.

Microsoft itself warns users that Copilot can make mistakes, but this incident shows how damaging those mistakes can be when AI output is treated as fact. This wasn’t a typo or a minor error. It was made-up information that influenced a policing decision.

The timing couldn’t be worse for Microsoft. With Copilot being integrated into nearly every part of Windows 11, to the point that it is frustrating users, incidents like this raise serious questions about trust, verification, and how much responsibility users and institutions should place on AI-generated content.