KPMG Allegedly Published AI Report Filled With Hallucinations

2026-06-15 · Source: Welcome to the Artificial Intelligence Incident Database · Field: Business & Management — Consulting & Professional Services, Corporate Strategy & Leadership · Depth: Fundamental Awareness, quick

Summary

KPMG, a major consulting firm, is accused of publishing a now-deleted AI report containing 40 out of 45 allegedly hallucinated citations. This report was subsequently referenced by OpenAI's ChatGPT, Google Gemini, a major Czech newspaper, and several trade publications. Specific examples include false claims about Japan's JR East using agentic AI, linking to a 2019 press release, and Austria's Verbund employing AI agents for real-time household analytics, despite no supporting evidence. AI safety firm GPTZero suggests an LLM research tool conflated sources and exaggerated claims, implying a lack of human verification. KPMG has removed the report and stated it is reviewing the circumstances, reiterating its guidelines on responsible AI use and human oversight. This incident mirrors a previous case where Deloitte refunded part of an AU\$440,000 report due to AI hallucinations.

Key takeaway

For Directors of AI/ML or consultants developing client-facing reports, you must implement stringent human oversight and verification protocols for all AI-generated content. Unverified AI outputs, especially citations, risk severe reputational damage and the propagation of misinformation through your work and downstream systems. Ensure your teams are trained to independently validate every claim and source, preventing "workslop" from undermining your credibility and client trust.

Key insights

Unverified AI-generated content, particularly citations, can lead to widespread misinformation and reputational damage.

Principles

• Human oversight is essential for validating AI-generated content.
• AI research tools can conflate sources and exaggerate claims.
• Unverified AI output can propagate through other LLMs and publications.

In practice

• Establish robust human review processes for AI-assisted reports.
• Independently verify all citations and factual claims from AI tools.
• Train teams on responsible AI use and content validation.

Topics

• AI Hallucinations
• Responsible AI
• Content Verification
• Consulting Ethics
• Large Language Models
• Misinformation

Best for: CTO, VP of Engineering/Data, AI Product Manager, Consultant, Director of AI/ML, AI Ethicist

Related on AIssential

• KPMG drops AI report after false case studies exposed — AI hallucinations pose a significant risk to the factual integrity of professional reports.
• Chasing the Hallucinations: KPMG's AI-Powered Attempt at "Redefining Excellence" — AI-generated "vibe citations" and claims are prevalent in professional reports, risking widespread misinformation.
• Every AI output is a hallucination — Large Language Models inherently produce fluent, confident outputs that are often wrong yet indistinguishable from truth, posing significant reliability issues.
• KPMG pulls report on AI usage due to apparent hallucinations — AI-generated content requires rigorous human oversight to prevent factual inaccuracies and reputational damage.
• Someone Made a Fake Wikipedia for AI Hallucinations, and It’s Basically the Internet Eating Itself — Halupedia demonstrates AI's capacity for elaborate, authoritative-sounding fabrication, highlighting the challenge of distinguishing real from AI-generated content.
• South Africa yanks AI policy after AI-assisted drafting invents citations — AI-generated content, especially citations, requires rigorous human verification to maintain integrity.

See Counsel's argued verdicts on the open AI decisions leaders are weighing →

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Welcome to the Artificial Intelligence Incident Database.