KPMG pulls – KPMG removed a report published in October 2025 after organizations said its claims about their AI usage were untrue or misleading, with a research group pointing to inaccuracies linked to AI hallucinations.
A KPMG report about “agentic AI” didn’t survive contact with reality.
The professional services firm pulled a publication titled “Redefining excellence in the age of agentic AI” after a number of organizations said its claims about their own AI usage were untrue. The report had been published in October 2025, and the disputes surfaced soon after.
Research group GPTZero identified multiple inaccuracies in the report. GPTZero told the FT that the errors stemmed from AI hallucinations—meaning the document’s wording and assertions about other companies’ AI behavior appear to have been produced with the help of AI. then failed to match what those organizations said was happening.
The pushback wasn’t limited to one corner of industry. UBS, the UK’s National Health Service, Swiss Federal Railways, and Transport for London all told the FT that the report’s claims about their AI usage were either untrue or misleading.
KPMG says it removed the report while investigating. A KPMG spokesperson said the firm removed the report from its websites and is reviewing what happened. The spokesperson also emphasized KPMG’s AI expectations for its own people. saying: “We expect all our people to follow our guidelines on the responsible use of AI. including human oversight to validate content and verify independent sources.”.
The incident lands in a moment when trust in AI-assisted writing is being stress-tested—especially when the subject is AI itself. KPMG’s pullout follows a similar case last month: EY withdrew a report on loyalty rewards programs that appeared to include fake footnotes and AI hallucinations.
The sequence is difficult to miss: a report about AI usage. claims about specific organizations. and then denials from those organizations paired with a finding that the inaccuracies may be driven by hallucinations. What starts as “agentic” ambition ends as a credibility problem—and for the companies named. it becomes an unwanted public correction.
KPMG agentic AI hallucinations GPTZero artificial intelligence cybersecurity AI reports responsible AI UBS NHS Swiss Federal Railways Transport for London EY
So basically KPMG lied and then yanked it. AI “hallucinations” is just a fancy way of saying they made stuff up.
Not surprised. Everyone uses AI to write faster now and then they’re shocked when it’s wrong. Also UBS and NHS saying it’s misleading? Yikes.
Wait, I thought “agentic AI” meant the AI is like… acting on its own, like robots doing audits? If it’s just hallucinating then what exactly were they even testing lol.
KPMG pulled it during an investigation sounds like PR to me. If other companies said the claims weren’t true, then why didn’t they check first? I saw an EY thing too about fake footnotes, so it’s like nobody can verify anything anymore. Next they’ll blame the interns or “prompting” or whatever.