Norwegian Man Files Complaint Against ChatGPT for False and Defamatory Claims

A Norwegian man, Arve Hjalmar Holmen, has filed a complaint against OpenAI after ChatGPT falsely claimed that he had killed his two sons and served a 21-year prison sentence. The chatbot’s inaccurate response came when Holmen searched for his own name. The digital rights organization Noyb is supporting his case, arguing that the AI-generated falsehood is defamatory and violates European data protection laws regarding personal data accuracy.

Holmen expressed concern that people might believe the fabricated information, highlighting the potential dangers of AI ‘hallucinations’—a known issue where AI models generate misleading or entirely false statements. OpenAI has been contacted for comment. Noyb criticized OpenAI’s approach, stating that a simple disclaimer about possible errors does not justify the spread of false information.

AI hallucinations remain a significant challenge for developers, with tech giants like Apple and Google also facing similar issues. Experts acknowledge that even AI researchers struggle to fully understand why these hallucinations occur. Since Holmen’s search in August 2024, ChatGPT has been updated to include real-time news sources, but concerns persist regarding the lack of transparency and data access in large language models.