The decade's most triggering comedy
In a statement Friday, the Guarantor for the Protection of Personal Data (GPDP), which oversees data privacy online, banned the U.S.-based chat website and its parent company, artificial intelligence developer OpenAI, from processing data from users in Italy. The agency said that OpenAI has no legal basis to collect data from Italian users to train the model, and has no age verification system to protect children against inappropriate answers.
“[There is] no way for ChatGPT to continue processing data in breach of privacy laws,” the agency said in a press release Friday. “The Italian [Supervisory Authority] imposed an immediate temporary limitation on the processing of Italian users’ data by OpenAI, the US-based company developing and managing the platform. An inquiry into the facts of the case was initiated as well.”
Much of the agency’s concern stemmed from the amount of information the app collects to train its language models. “[N]o information is provided to users, nor to interested parties whose data was collected by OpenAI, LLC and processed through the ChatGPT service,” the GPDP order stated. The order also noted that the data processed by ChatGPT can be inaccurate, since the AI does not always match factual circumstances.
Furthermore, the order states that there is an “absence of a suitable legal basis in relation to the collection of personal data and their treatment for the purpose of training the algorithms underlying the functioning of ChatGPT.” The agency also cited a data breach that occurred on March 20, revealing both conversations and payment information of some who use the more sophisticated, paid-for version of ChatGPT.
The other issue highlighted by the Italian government was worries about age verification. OpenAI’s terms of service say that the service is for users over the age of 13, but Italy said the lack of age verification filters for minors “exposes them to absolutely unsuitable responses with respect to their degree of development and self-awareness.” The agency said that because of the lack of age verification and the complex violations of Italian law, it had no choice but to impose a blanket ban on harvesting data from all users.
The ban went into effect immediately. OpenAI must notify the GPDP of what measures it is taking to comply with the order within 20 days, otherwise, the country will impose a fine of up to 20 million euros or 4% of the company’s global revenue.
The move from the Italian government comes just days after a Belgian man reportedly committed suicide after interacting with a different chatbot. According to Euronews, a man in his 30s, with a wife and two young children, committed suicide after interacting for several weeks with ELIZA, another chatbot powered by a different language model.
According to the man’s wife, who spoke with several European news outlets, the man became fixated on climate change and spent his days confiding his fears in ELIZA. The chatbot allegedly stoked the man’s fears until he developed suicidal ideation; the man reportedly suggested that he sacrifice himself so that ELIZA could save humanity through AI, his widow said.