West Midlands Police chief constable Craig Guildford, the head of the UK police force that admitted earlier this week that it had used faulty information from Microsoft’s Copilot AI assistant in its controversial decision to ban Israeli football fans from a November 2025 Europa league match, has announced his resignation (via the BBC). His retirement follows days of criticism over the scandal, with UK home secretary Shabana Mahmood and Downing Street both expressing that they had “lost confidence” in the chief constable.
Rather than his force’s failure to fact-check a notoriously unreliable technology, Guildford said in a statement announcing his retirement that “the political and media frenzy” was responsible for his decision to resign.
“The political and media frenzy around myself and my position has become detrimental to all the great work undertaken by my officers and staff in serving communities across the West Midlands,” Guildford said. “I have carefully considered my position and concluded that retirement is in the best interests of the organisation, myself and my family. It has been the honour of my career serving as the Chief Constable of West Midlands Police.”
In his statement, Guildford did not apologize for the West Midlands Police’s handling of the Aston Villa v Maccabi Tel Aviv match, or for its inclusion of factually inaccurate information in the preceding intelligence report.
In November 2025, the Birmingham Safety Advisory Group—led by West Midlands Police and Birmingham City Council—made the contentious decision to ban Israeli fans from attending an upcoming Europa League match between Aston Villa and Maccabi Tel Aviv, fearing a repeat of November 2024’s violent clashes over Gaza war tensions with Israeli football fans in Amsterdam.
That decision was informed by an intelligence report submitted by the West Midlands Police, which—in addition to featuring interpretations of the 2024 Amsterdam fan violence that have since been questioned by Dutch authorities—referenced an earlier West Ham v Maccabi Tel Aviv match in 2023. That West Ham v Maccabi Tel Aviv match, however, never occurred, which is the exact kind of error you don’t want in an intelligence report.
In a letter to the Home Affairs Committee earlier this week, Guildford admitted the West Ham match was fabricated by Microsoft’s Copilot AI assistant, which the WMP had consulted and failed to fact-check while compiling its report.
“On Friday afternoon I became aware that the erroneous result concerning the West Ham v Maccabi Tel Aviv match arose as result of a use of Microsoft Copilot,” Guildford said. “Both ACC O’Hara and I had, up until Friday afternoon, understood that the West Ham match had only been identified through the use of Google.”
In December, Guildford had been questioned in a Home Affairs Committee hearing investigating the handling of the Aston Villa v Maccabi Tel Aviv match. When asked by a minister of parliament whether the WMP “did an AI search, got something about West Ham, and just whacked it into” the report, Guildford said “No, not at all,” claiming he instead believed the nonexistent football match to have emerged from the police force’s “search through social media.”
In his letter, Guildford acknowledged that the WMP had indeed just whacked the results of an AI search into its intelligence report, and offered his “profound apology to the Committee for this error.”
Like other LLM-driven assistants, Copilot features a disclaimer that it “may make mistakes,” which you might hope would encourage the personnel of security, law enforcement, and defense organizations who are utilizing LLM technologies to exercise due caution. But when those technologies, despite their infamous unreliability, are marketed as one-button replacements for human knowledge and problem-solving, it’s hard to think of these types of oversights as being anything but inevitable.
In November, Microsoft AI CEO Mustafa Suleyman said it’s “mindblowing” that people aren’t more impressed with generative AI. Mustafa, if you’re listening: This is why. Thankfully, it’s not like we’re setting ourselves up for even higher stakes AI embarrassments.
In other news, US secretary of defense Pete Hegseth announced this week that the Pentagon will begin integrating X’s Grok AI into military networks. Yes, the same LLM platform drawing growing outrage for flooding X with generated pornographic imagery of nonconsenting people, including minors. Surely nothing could go wrong with that, either.
