# KPMG Pulls Agentic AI Report: Invented Case Studies Laundered Through Trusted Logos

Times of AI Desk · 2026-06-13 · Industry

[https://timesof.ai/2026/06/kpmg-pulls-agentic-ai-report-alleged-hallucinations](https://timesof.ai/2026/06/kpmg-pulls-agentic-ai-report-alleged-hallucinations)

> KPMG removed “Redefining excellence in the age of agentic AI” after UBS, NHS, SBB, TfL and others disputed fabricated adoption stories. GPTZero/FT pointed to hallucinations—AI-written marketing about AI success until the subjects objected.

When a global consulting brand publishes invented AI case studies, it damages the evidence base enterprises use to justify agentic AI spend. AI-written marketing about AI success, laundered through trusted logos, can move markets and client decisions until the subjects themselves object.

Big Four firm **KPMG** pulled its agentic-AI thought-leadership report after a cascade of denials from organizations featured as case studies. The October 2025 paper, *Redefining excellence in the age of agentic AI*, had touted enterprise AI adoption stories that UBS, the NHS, Swiss Federal Railways, Transport for London, and others told the Financial Times were **false or misleading**. Detection work by GPTZero and FT reporting pointed to **AI-generated hallucinations**—a professional-services firm apparently using AI to write an AI excellence report.

## What happened

Per TechCrunch, FT-cited coverage, and KPMG’s public response:

- **Report**: “Redefining excellence in the age of agentic AI” (published ~October 2025).
- **Action**: Removed from KPMG websites while the firm conducts an internal investigation.
- **Complaints**: Named organizations said descriptions of their AI programs did not match reality.
- **Organizations cited in coverage**: **UBS**, **U.K. National Health Service**, **Swiss Federal Railways**, **Transport for London**, among others.
- **Alleged cause**: AI hallucinations in drafting—GPTZero flagged inaccuracies consistent with generative fabrication rather than ordinary editorial error.

The episode became a weekend-to-Monday industry story alongside continued fallout from Anthropic’s June 12 Fable/Mythos suspension (covered separately).

## Limits

- Hallucination assessment is GPTZero/FT-cited analysis—not a completed forensic audit published by KPMG.
- Firm is investigating; final cause determination was not public on pull day.
- Named orgs disputed content; full original report text is no longer on KPMG sites for independent re-check.

## Sources

- TechCrunch: “KPMG pulls report on AI usage due to apparent hallucinations” (June 13, 2026).
- Financial Times reporting cited by TechCrunch/GPTZero on organization denials and hallucination assessment (June 2026).
- KPMG spokesperson statement on removal pending investigation (via contemporaneous coverage).