KPMG pulls agentic AI report over hallucinations

The rush by major consultancies to position themselves as AI experts has resulted in a lazy, unverified reliance on the very technology they are analyzing, leading to embarrassing reputational damage. When professional services firms bypass basic human verification to pump out thought leadership, they compromise the integrity of corporate research.

• –The incident demonstrates that "vibe citing"—where LLMs generate plausible but entirely fake references—is slipping past professional editorial teams.
• –Rushing to publish AI-themed reports has led both KPMG and EY to commit the same error of using unverified generative AI, suggesting systemic pressure to produce high volumes of AI content.
• –AI detection tools like GPTZero are expanding their role from academic plagiarism detectors to corporate auditing tools capable of uncovering corporate research negligence.