Italian data protection authorities have decided to temporarily block ChatGPT nationwide due to allegations that OpenAI has been processing unlawful personal data. The basis for the allegations is a recent data leak that allowed access to information of other users, which included payment data. The users and persons whose data were collected by OpenAI were allegedly not informed, and there was no legal basis for massive collection and storage of personal data for the purpose of training algorithms.
OpenAI’s CEO, Sam Altman, stated that there was a bug in an open-source library, but failed to provide specific details. The use of user input to train AI is also said to be problematic. Moreover, checks revealed discrepancies between the information provided by ChatGPT and actual data.
OpenAI releases ChatGPT for people above the age of 13, but does not verify users’ ages in the web app. Hence, young people may receive inappropriate responses that are not suitable for their age. Missing age checks are common, and most social networks do not verify users’ ages.
The Italian data protection supervisory authority has ordered the ChatGPT block at OpenAI. However, the web app could still be accessed via a VPN tunnel with an Italian IP. To avoid risking fines, geo-blocking by OpenAI may be implemented. Meanwhile, further investigations have been launched.
