ChatGPT Falsely Accused Australian Mayor of Bribery, Prompting First AI Defamation Lawsuit Threat

In April 2023, ChatGPT generated false information about Brian Hood, the mayor of Hepburn Shire in Victoria, Australia, incorrectly stating that he had been convicted of bribery in connection with a foreign corruption scandal involving the Reserve Bank of Australia's subsidiaries Securency and Note Printing Australia. The AI chatbot's response fundamentally misrepresented Hood's actual role in the scandal, which was as a key whistleblower who helped expose the corruption rather than as a perpetrator. The real facts of the case show that Hood was working as an executive at the Reserve Bank of Australia's Note Printing Australia subsidiary when he discovered evidence of bribery involving payments to foreign officials to secure currency printing contracts. Rather than participating in the corruption, Hood courageously reported the illegal activities to authorities and played a crucial role in exposing what became known as one of Australia's largest corporate bribery scandals. Several other executives were ultimately prosecuted and convicted, but Hood was never charged with any wrongdoing. When confronted with the AI-generated defamatory content, Hood's legal team sent a concerns notice to OpenAI, the company behind ChatGPT, demanding correction of the false information. This action represented what would have been the first defamation lawsuit filed against an AI chatbot anywhere in the world. The notice cited the significant reputational damage caused by the false accusations, particularly given Hood's public role as an elected official and his actual status as a whistleblower in the case. The incident highlighted critical issues around AI systems generating false information about real individuals, particularly when such information involves serious criminal allegations. Legal experts noted the complexity of pursuing defamation claims against AI systems, including questions about liability, jurisdiction, and the technical challenges of ensuring AI models don't repeat false information. The case also raised broader concerns about the reliability of AI-generated content and the need for better safeguards when AI systems discuss real people and legal matters. OpenAI ultimately addressed the concerns, though specific details of the resolution were not publicly disclosed. The incident served as a wake-up call for the AI industry about the legal risks of AI hallucinations involving real individuals and the potential for significant reputational harm when AI systems confidently state false information about people's criminal histories or legal troubles.