KPMG Retracts AI Report After Clients Dispute Fabricated Claims

KPMG has removed a report on enterprise artificial intelligence adoption from its websites after several prominent organizations publicly contradicted the document's claims about their technology deployments.

The report, titled "Redefining excellence in the age of agentic AI" and published in October 2025, contained statements about AI usage at major institutions that those organizations say were either false or misleading. UBS, the United Kingdom's National Health Service, Swiss Federal Railways, and Transport for London all told the Financial Times that KPMG's characterizations of their AI initiatives were inaccurate.

Research organization GPTZero, which analyzes AI-generated content, identified multiple inaccuracies in the KPMG document and attributed them to AI hallucinations—a phenomenon where generative AI systems produce plausible-sounding but factually incorrect information. The assessment suggests KPMG may have used AI tools to help produce a report ostensibly analyzing AI adoption trends.

Why it matters

The incident exposes a critical vulnerability in professional services firms' own use of the AI tools they recommend to clients. When consulting giants publish research containing fabricated claims about real organizations, it undermines trust in both their advisory work and the broader enterprise AI market. The episode also demonstrates how AI hallucinations can slip past quality controls even at firms with extensive review processes, raising questions about verification standards across the industry.

A pattern emerges

KPMG's withdrawal follows a similar incident last month involving competitor EY, which retracted a report on customer loyalty programs after the document was found to contain fake footnotes and apparent AI hallucinations. The back-to-back retractions at two of the world's largest professional services firms suggest quality assurance processes have not kept pace with the rapid internal adoption of generative AI tools.

A KPMG spokesperson acknowledged the firm is conducting its own investigation into how the inaccuracies occurred. "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.

Verification challenges

The episode highlights the difficulty of detecting AI-generated errors, particularly when they involve specific claims about real organizations. Unlike obvious fabrications, statements about corporate AI deployments can sound plausible to editors unfamiliar with each client's actual technology stack. Without direct verification with the organizations mentioned, such claims can pass through multiple review stages.

For enterprises evaluating AI tools, the KPMG incident serves as a reminder that even authoritative-seeming research requires scrutiny. Organizations mentioned in third-party reports may want to establish processes for monitoring how their technology initiatives are characterized in published materials.

The details were first reported by the Financial Times.