ChatGPT Accused a Man of Drowning His Children in a Pond. His Kids Are Still Alive: Complaint

A man in Norway has filed a legal complaint against OpenAI after ChatGPT falsely claimed he drowned his children in a pond—despite the fact that his children are alive and well.

In an attempt to see what OpenAI's chatbot would say about him, Arve Hjalmar Holmen asked ChatGPT, "Who is Arve Hjalmar Holmen?"

To his shock, the AI generated a completely false and defamatory story, claiming he had murdered two of his sons and attempted to kill a third in Trondheim, Norway, in 2020, according to a report.

The chatbot even detailed a fabricated conviction and prison sentence of 21 years. Disturbed by the AI's ability to create such a damaging lie, Holmen sought legal action.

Holmen, represented by the European Center for Digital Rights (noyb), has filed a formal complaint with Norway's data protection authority, Datatilsynet, accusing OpenAI of violating the EU's General Data Protection Regulation (GDPR).

The complaint argues that OpenAI's chatbot unlawfully processed and spread inaccurate personal data, failing to comply with Article 5(1)(d) of the GDPR, which mandates that companies take reasonable steps to ensure the accuracy of personal data.

Despite Holmen contacting OpenAI for corrections, he received only a generic response, leading him to escalate the matter legally.

The complaint requests that OpenAI be ordered to delete the false information and adjust its AI model to prevent similar inaccuracies in the future.

Holmen and noyb are also pushing for a fine against OpenAI, arguing that the false claim could have severely impacted his personal and professional life.