A Norwegian man named Arve Hjalmar Holmen is upset because ChatGPT, a chatbot, wrongly said that he killed his two sons and was sent to prison for 21 years. This false claim hurt Holmen a lot. He found out about the lie when he asked ChatGPT, “Who is Arve Hjalmar Holmen?” The chatbot replied with a made-up story that said his boys, aged 7 and 10, were found dead in a pond in Trondheim, Norway, in December 2020. Holmen insists that this isn’t true and it scared him that people might believe such false information.
Holmen has made a complaint to the Norwegian Data Protection Authority, asking for OpenAI, the company that makes ChatGPT, to be punished for the wrong information. He said, “People might think there’s no smoke without fire,” which means they might believe something is true just because they heard it.
OpenAI has admitted the mistake and said it happened in an older version of ChatGPT. They mentioned that they are working on ways to make their chatbot better at providing accurate information and to prevent such mistakes in the future.
A digital rights group called Noyb is supporting Holmen’s complaint. They argued that what ChatGPT said is not only wrong but also breaks European laws about keeping personal information accurate. Their complaint states that “Holmen has never been accused or convicted of any crime and is a good citizen.”
Even though ChatGPT has a warning saying it can make mistakes, Noyb believes that this is not enough. A lawyer from Noyb, Joakim Söderberg, said you can’t just share false information and think a simple warning is okay.
This story highlights the bigger problem of AI spreading false information. Recently, Apple stopped its news summary tool in the UK after it shared wrong headlines, and Google’s AI also made silly suggestions, like sticking cheese on pizza with glue
Leave a Reply