KPMG's AI Report Becomes Accidental Demo of AI Hallucinations
Key Takeaways
- ▸Only 5 of 45 citations in KPMG's report correctly pointed to their sources; the rest were mangled, misleading, or fabricated
- ▸Roughly 50% of the report's factual claims were false or unsupported, including false case studies about corporate AI deployments
- ▸The report contradicted KPMG's own internal research—claiming 55% of CEOs ranked AI as top priority when KPMG's own 2025 CEO Outlook reported 71%
Summary
KPMG's October 2025 report titled 'Total Experience: Redefining Excellence in the Age of Agentic AI' has become a cautionary tale about AI hallucinations after research firm GPTZero found that only 5 of its 45 citations correctly referenced their sources. GPTZero's forensic review revealed that roughly half of the report's factual claims were false, unsupported, or misattributed—a phenomenon GPTZero termed 'vibe citing.' The report contained multiple false case studies about agentic AI deployments at major organizations including UBS, Swiss Federal Railways, and Transport for London, as well as fabricated claims about Emirates' Sara chatbot. The irony is sharp: KPMG has long advised clients on managing precisely these AI risks. The consulting firm has since removed the report and launched an investigation, with a spokesperson emphasizing the need for human oversight and source verification in AI-generated content.
- KPMG is investigating how an AI-generated report with such widespread errors passed editorial review and was published
Editorial Opinion
This incident exposes the growing gap between Big Four consulting firms' confident AI advice and their actual deployment practices. While KPMG and peers like Deloitte (which refunded the Australian government for AI-generated content) counsel clients on managing AI risks, they're simultaneously demonstrating why those warnings exist. 'Vibe citing'—where AI stitches together fragments that look credible—is exactly the kind of hallucination these firms promise to help clients avoid. The lack of human oversight in KPMG's process undermines their credibility as AI governance advisors.
