KPMG's AI-written report praising AI withdrawn after fake cases surface

Artificial intelligence (AI) is supposed to help businesses work faster and smarter. But what happens when AI starts inventing facts — and those facts end up in reports published by some of the world's biggest consulting firms?

That question is at the centre of an embarrassing controversy involving KPMG, one of the world's largest professional services firms, after it was forced to withdraw a major global report that contained several false claims and fabricated case studies about how leading organisations were supposedly using AI.

The episode has reignited concerns about AI "hallucinations" and whether companies are moving too quickly to embrace the technology without proper fact-checking.

According to a report by the Financial Times, the withdrawn publication included examples of AI adoption that several organisations later said were inaccurate, misleading or entirely false.

WHAT WAS IN THE KPMG REPORT?

The controversy centres around a report titled "Redefining Excellence in the Age of Agentic AI", which highlighted how companies were supposedly using advanced AI systems to transform their operations.

However, according to the report, several of the case studies cited by KPMG never actually happened. The examples appeared to be the result of AI-generated hallucinations — a term used when artificial intelligence systems confidently generate information that sounds convincing but is not true.

The issues were first flagged by AI detection and research firm GPTZero and were later verified by the Financial Times. Following the findings, several organisations named in the report reportedly demanded that KPMG remove the publication.

BIG ORGANISATIONS SAID THE CLAIMS WERE FALSE

Among the organisations that challenged KPMG's claims were Swiss banking giant UBS, the UK's National Health Service (NHS), Swiss Federal Railways and Transport for London.

KPMG had claimed that UBS was using sophisticated AI agents across investment advisory and risk management systems through a Microsoft-built platform. UBS rejected the claims and described them as "factually incorrect", according to the report.

The report also claimed that Swiss Federal Railways was using AI agents to help passengers plan and book journeys using real-time travel and carbon-emissions data. Railway officials reportedly responded by saying the claims were "not accurate".

Transport for London was another organisation cited in the report. KPMG claimed the transport authority was using AI agents to predict congestion and coordinate transport services across the city. TfL reportedly described those claims as "misleading".

Perhaps the most striking example involved NHS Greater Manchester. The report suggested that AI agents were organising patient records, automating referrals and predicting hospital readmissions. Health officials reportedly responded by saying the description "doesn't really align" with reality.

A BIGGER PROBLEM FOR THE AI INDUSTRY?

The incident has raised broader questions about how companies are using generative AI tools in research, consulting and publishing.

This is not the first such controversy involving a major consulting firm.

According to the report, EY was recently forced to withdraw a separate study after GPTZero identified fake footnotes and AI-generated errors in the document.

GPTZero Chief Executive Edward Tian warned that mistakes made by trusted organisations can spread quickly and be repeated across the internet before they are corrected.

"They poison the well of information," Tian said, noting that some of KPMG's claims had already been cited by technology publications and even a major European newspaper before the report was removed.

The concern is particularly important as businesses increasingly rely on AI-generated content for reports, presentations, research papers and internal decision-making.

KPMG RESPONDS

Following the controversy, KPMG International removed the report and launched an internal investigation.

A spokesperson said the firm takes the "accuracy and integrity" of its publications seriously and acknowledged that employees may have breached the company's internal policies on AI usage.

"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 company is now reviewing how the report was produced and whether internal controls were followed.

The KPMG episode highlights one of the biggest challenges facing the AI industry today.

While artificial intelligence can dramatically speed up research and content creation, it can also generate false information that appears credible unless checked carefully by humans.

For consulting firms, auditors and professional services companies whose business depends on trust and accuracy, such mistakes can carry significant reputational risks.

The controversy is also a reminder that as AI adoption accelerates, human oversight remains critical. A convincing report may look polished and professional, but if the underlying facts are wrong, the consequences can be costly.

As businesses race to integrate AI into everyday work, the KPMG incident may become one of the most prominent examples of why speed cannot replace verification.