Big Four consultancy PwC is reassessing four recent “thought-leadership” reports after an investigation found multiple AI-generated hallucinations and errors. The case highlights the need for human oversight of AI-written content and for AI governance to be integrated with compliance programs.
While the case has proved embarrassing for PwC, the stakes could be far higher for compliance teams that rely unquestioningly on AI tools when reporting to stakeholders or regulators. False attributions or repeated inaccuracies could lead to fines and undermine investor trust.
