KPMG withdraws agentic AI report after disputed case studies

KPMG has pulled a global report on agentic AI after multiple organisations challenged case studies that described how they were using AI agents. The report, titled Redefining Excellence in the Age of Agentic AI, said companies and public bodies were applying AI agents across finance, healthcare and transport, but several examples were found to be inaccurate and appeared to stem from AI hallucinations.

GPTZero first flagged the errors, which were later verified by the Financial Times. UBS disputed a claim that it used AI agents in investment advisory and risk management through a custom Microsoft-built platform, calling it “factually incorrect”. Swiss Federal Railways said a description of AI agents helping passengers plan and book journeys based on real-time conditions and carbon emissions was “not accurate”.

Transport for London said claims about using AI agents to predict congestion and coordinate the city’s transport network were “misleading”. NHS Greater Manchester also challenged KPMG’s description of AI agents organising patient records, automating referrals and predicting hospital readmissions, saying it “doesn’t really align” with reality.

KPMG International has removed the report and started an internal investigation. A spokesperson said the firm takes the “accuracy and integrity” of its publications seriously and expects employees to follow responsible AI guidelines, including human oversight and source verification. The incident follows EY’s withdrawal of a study that reportedly contained fake footnotes and AI-generated errors.