KPMG Pulls AI Report After Hallucination Claims

News

1. TCS Ties Up With Anthropic to Push Claude Into Enterprise AI

2. Karnataka Shortlists 256 Deeptech Startups for ELEVATE NxT Finale

3. Reliance-Backed Addverb Eyes $100 Million for Robotics Push

4. Anthropic Releases Claude Fable 5 and Mythos 5

5. Meta to Lease First AI Data Center in India From Reliance

6. Musk Says Space AI Data Centers Are Within Reach as SpaceX Nears IPO

KPMG has withdrawn a report on agentic artificial intelligence after researchers and the Financial Times found that it contained inaccurate claims and apparently fabricated or distorted citations, in the latest embarrassment for a professional services firm promoting AI adoption while struggling with its own verification standards.

The report, titled Total Experience: Redefining Excellence in the Age of Agentic AI, was published in October 2025 and presented examples of how businesses and public bodies were using AI.

According to the FT, several organizations named in the report, including UBS, the UK’s National Health Service, Swiss Federal Railways and Transport for London, said the descriptions of their AI use were inaccurate or misleading.

The report claimed that UBS “integrates AI agents across investment advisory, risk management and compliance monitoring” through a platform co-developed with Microsoft.

A UBS spokesperson told the FT the assertions were “factually incorrect”.

KPMG also said Swiss Federal Railways offered AI agents that helped users plan, book and optimize journeys based on preferences, real-time conditions and carbon impact. A spokesperson for the railway said the assertion was “not accurate”.

The report further claimed that Transport for London was using AI agents “to predict and manage congestion, personalize commuter updates and co-ordinate multimodal transport”. A spokesperson for the transit system told the FT the assertion was “misleading”.

Another claim, that NHS Greater Manchester used AI agents to predict hospital readmissions, triage patients and automate referrals, “doesn’t really align” with the press release on which it appeared to be based, a spokesperson said. The FT reported that the cited release was about an AI tool designed to combat lung cancer and did not mention agentic AI carrying out those tasks.

Research group GPTZero said it reviewed 45 citations in the report and found that only five accurately pointed to real sources.

It said 28 citations included paraphrased titles or fake components for real sources, while 12 were too vague or flawed to verify confidently.

GPTZero described the pattern as “vibe citing,” where AI-generated references appear plausible but do not accurately correspond to the underlying sources.

The group said several claims appeared to be fake or misattributed, likely because an AI research tool over-complied with a request to find examples of agentic AI in use.

KPMG removed the report from some of its websites after being alerted to the issues.

A company spokesperson told The Register that KPMG International “takes the accuracy and integrity of its published content seriously” and was reviewing the circumstances around the publication.

“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, according to the FT.

The episode is striking because the disputed report was about agentic AI, a class of systems designed to perform tasks with greater autonomy.

Consulting firms have been aggressively advising clients on AI adoption, governance and risk controls, while also using the same tools internally to produce research and marketing material.

The KPMG case follows a similar incident involving EY, which withdrew a report last month after GPTZero identified fake footnotes and other errors.

Deloitte also faced scrutiny in Australia after AI-generated errors appeared in a government-commissioned report, prompting the firm to refund part of its payment.

The issue is not confined to consulting. AI-generated citations and false references have surfaced in academic papers, legal filings and government reports, raising questions about how institutions verify AI-assisted work before publication.

For the Big Four firms, the risk is sharper because their research is often cited by companies, policymakers and media outlets as authoritative. If such reports contain false claims, the errors can travel quickly into presentations, boardroom decisions and further AI-generated summaries, analysts said.