A Norwegian father is in a legal battle with OpenAI after ChatGPT falsely accused him of murdering his children. The AI chatbot fabricated a story claiming he was sentenced to 21 years in prison for the crime.
Arve Hjalmar Holmen was shocked by ChatGPT’s false accusation. The chatbot mixed real details like the number and gender of his children with the fabricated murder story.
Holmen, with the help of digital rights group Noyb, filed a complaint alleging GDPR violations. He argues that the false information harms his reputation and is difficult to correct.
“Some think that ‘there is no smoke without fire,’” Holmen stated, expressing his concern that people might believe the false story.
While ChatGPT has been updated to prevent repeating the false claim, Noyb argues the inaccurate data likely remains in its internal dataset. They demand OpenAI delete the defamatory output and refine its model.
This isn’t an isolated incident. Previously, ChatGPT has been known to generate false information, leading to defamation concerns. An Australian mayor threatened to sue after the chatbot falsely claimed he went to prison.
Noyb argues that disclaimers are insufficient. “AI companies should stop acting as if the GDPR does not apply to them,” said Noyb data protection lawyer, Kleanthi Sardeli.
Noyb has filed a complaint with Norwegian data authorities, seeking an order requiring OpenAI to delete the defamatory output and fine-tune its model to eliminate inaccurate results.
OpenAI faced a $16 million fine and a temporary ban in Italy for data breaches. To restore ChatGPT, OpenAI was ordered to allow users to correct their personal data.
If Norwegian authorities rule against OpenAI, the company may need to overhaul ChatGPT’s algorithm.
Source: Ars Technica