KPMG pulls report on AI usage due to apparent hallucinations

KPMG, the prominent professional services firm, has decided to retract a report titled "Redefining excellence in the age of agentic AI" following claims from several organizations that its assertions regarding their AI implementations were inaccurate. The research group GPTZero uncovered multiple discrepancies within the report, which was released in October 2025. GPTZero indicated that these inaccuracies were the result of what is known as AI hallucinations, a phenomenon where AI generates incorrect or misleading information. This raises concerns about the reliance on AI-generated content for such critical reports. Notably, organizations including UBS, the UK’s National Health Service, Swiss Federal Railways, and Transport for London have all stated that the report misrepresented their AI usage. In response to the backlash, a KPMG representative confirmed that the report has been removed from their online platforms as the firm conducts its own review of the situation. The spokesperson emphasized the importance of adhering to guidelines regarding the responsible use of AI, which includes ensuring human oversight to validate information and verify sources. This incident follows a similar event last month when EY retracted a report on loyalty rewards programs due to the inclusion of fabricated footnotes and errors attributed to AI hallucinations.