Incident Overview
On November 20, 2025, West Midlands Police used an intelligence report generated by Microsoft Copilot to justify the banning of Maccabi Tel Aviv football supporters from attending a Europa League match against Aston Villa in Birmingham. The report contained fabricated claims, including references to events that had never occurred. This decision directly impacted Maccabi Tel Aviv supporters, the football club, the Jewish community in Birmingham and the West Midlands, and the broader issue of epistemic integrity in policing.
The ban was imposed based on the erroneous intelligence, which cited a non-existent incident. The force later admitted to MPs that the information had been generated by Copilot and was factually incorrect. This incident marks one of the first documented cases where AI-generated content directly influenced a law enforcement decision with tangible consequences for a specific group.
The error was acknowledged by Chief Constable Craig Guildford, who apologized to MPs for providing incorrect evidence. The admission underscores the risks of relying on AI systems in high-stakes decision-making processes, particularly in law enforcement where accuracy and fairness are paramount.
This case highlights the potential for AI systems to propagate misinformation, reinforce biases, and erode public trust when deployed without rigorous oversight or human verification. The incident also raises questions about the accountability of both the police force and the AI provider in such scenarios.
Possible Root Cause
The primary root cause of this incident was the uncritical adoption of AI-generated content by West Midlands Police without adequate fact-checking or human review. Copilot, an AI-powered tool designed to assist with drafting and research, produced a report that included fabricated events. These inaccuracies were not detected before the report was used to justify the ban on Maccabi Tel Aviv supporters.
The reliance on AI-generated intelligence in policing decisions reflects a broader trend of integrating AI tools into operational workflows without sufficient safeguards. In this case, the police force did not implement protocols to verify the accuracy of AI outputs, leading to the dissemination of false information.
While Copilot is designed to aid users, its outputs are not inherently trustworthy without human validation. The police force’s failure to cross-reference the AI-generated report with reliable sources or independent verification mechanisms contributed to the error.
The speed at which AI can generate content can create a false sense of reliability, encouraging users to accept outputs at face value. This case serves as a cautionary example of the dangers of prioritizing speed over accuracy in critical decision-making processes.
Harm and Impact
- Misinformation and Fabricated Claims: The Copilot-generated report included references to events that had never taken place, leading to the unjustified banning of Maccabi Tel Aviv supporters. This misinformation not only affected the individuals targeted but also undermined the credibility of the police force and the AI tool itself.
- While the intent behind the ban may not have been discriminatory, the outcome had the potential to alienate and stigmatize a specific group, contributing to broader societal tensions.
- Erosion of Trust and Democratic Harm: The incident eroded public trust in both the police force and AI systems used in law enforcement. When AI tools are deployed without transparency or accountability, they can undermine democratic principles by enabling decisions that lack a solid evidentiary basis. The admission of the error by Chief Constable Guildford was a step toward restoring trust, but the long-term reputational damage remains.
- Epistemic Integrity: The use of unverified AI-generated intelligence in a policing context threatens the foundational principle of epistemic integrity—the reliability and accuracy of information used in decision-making. This incident demonstrates how AI systems, when not properly vetted, can introduce false narratives into critical processes, compromising the integrity of institutional actions.
It raises broader concerns about the ethical deployment of AI in public sector decision-making, particularly in areas where accuracy and fairness are non-negotiable. The case also underscores the need for clearer guidelines on the use of AI tools in law enforcement, including mandatory human oversight and verification protocols.
Sources
Incident summary from the AI Incident Database, used under CC BY-SA 4.0.