On Thursday, 20 March, the NGO NOYB lodged a complaint with the Norwegian data protection agency, Datatilsynet, accusing OpenAI and its artificial intelligence, ChatGPT, of breaching the [General] Data Protection Regulation (GDPR) due to a defamation matter.
A Norwegian ChatGPT user was unpleasantly surprised to see the chatbot indicate that he was guilty of having “murdered” two of his children when he asked it what information the AI tool had about him.
The story was completely false but had real information, including the city where the user was born and how many children he has. This error prompted the NGO to lodge a complaint naming the respondent: “By knowingly allowing ChatGPT to produce defamatory results, the company clearly violates the GDPR’s principle of data accuracy”, NOYB affirms in its press release.
The NGO also accuses OpenAI of “trying to get around” its legal responsibilities under the GDPR, starting with updating user information and deleting it if need be.
“You can’t bypass the legal obligation to ensure the accuracy of the personal data you process via a disclaimer”, insists NOYB.
The lack of certainty about whether the error will be completely rectified is also a problem for the association. “There is no way for the individual to be absolutely sure that this output can be completely erased [...], unless the entire AI model is retrained”, warns the NGO.
See the complaint: https://aeur.eu/f/g3p (Original version in French by Isalia Stieffatre)