Professional services firm KPMG has withdrawn a report on artificial intelligence after multiple organizations, including UBS and the UK’s National Health Service, said the report’s claims about their use of AI were false. The report, titled “Redefining excellence in the age of agentic AI,” was published in October 2025 and later flagged by research group GPTZero for containing numerous inaccuracies that GPTZero attributed to AI hallucinations.
GPTZero told the Financial Times that the errors appeared to stem from the use of AI to help write the report — an irony given the report’s focus on AI excellence. UBS, the NHS, Swiss Federal Railways, and Transport for London all told the FT that the report’s assertions about their AI deployments were either untrue or misleading.
Also read: Nous Research in talks to raise $75M+ at $1.5B valuation for open-source AI agent Hermes
Investigation underway
A KPMG spokesperson confirmed the firm removed the report from its websites while conducting an internal investigation. “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 spokesperson said.
The incident follows a similar case last month, when EY withdrew a report on loyalty rewards programs that appeared to contain fabricated footnotes and AI-generated hallucinations.
Also read: The wildest allegations in Apple’s trade secrets lawsuit against OpenAI
Industry implications
The back-to-back retractions by two of the Big Four accounting firms raise questions about how professional services organizations are deploying AI internally — and whether their own AI governance practices match the advice they sell to clients. KPMG, like its peers, has invested heavily in AI tools and markets AI consulting services to corporate clients.
GPTZero, which develops tools to detect AI-generated text, said the KPMG case underscores the need for stronger verification processes in corporate publishing. The group’s analysis found that several claims in the report could not be traced to any public source, suggesting the text was generated without proper fact-checking.

Be the first to comment