According to Italy’s data protection agency Garante, ChatGPT producer OpenAI plans to present measures on Thursday to address concerns that led to an Italian ban on the chatbot last week.
Microsoft Corporation-backed (MSFT.O) OpenAI took ChatGPT offline in Italy after Garante temporarily restricted it last week and launched an investigation into a possible violation of privacy rules.
The agency accused OpenAI last week of failing to verify the age of ChatGPT users and of having “no legal basis that justifies the massive collection and storage of personal data.”
On Thursday, it stated that it has no intention of halting AI development, but emphasized the importance of adhering to rules aimed at protecting the personal data of Italian and European citizens.
OpenAI pledged to be more transparent about how it handles user data and verifies the user’s age in a video conference late Wednesday, attended by CEO Sam Altman, Garante said.
On Thursday, the company said it would send Garante a document outlining the steps it would take in response to its requests.
The data authority stated that it would evaluate OpenAI’s proposals. According to a source familiar with the situation, determining the contents of the letter will most likely take several days.
The San Francisco-based OpenAI did not respond to a request for comment on the agency’s statement.
The company said in a blog post titled “Our approach to AI safety” on Thursday that it was working to develop “nuanced policies against behavior that represents a genuine risk to people.”
“We do not use data to sell our services, advertise, or build people profiles,” it stated. “We use data to improve the utility of our models for people.” ChatGPT, for example, improves through additional training on the conversations people have with it.
“While some of our training data contains personal information that is freely available on the internet, we want our models to learn about the world, not individual people.”
According to the company, it has removed personal information from its datasets where possible, fine-tuned models to reject user prompts requesting such information, and will respond to individual requests to delete their data from its systems.
The Italian ban has piqued the interest of other European privacy regulators, who are investigating whether harsher measures are required for chatbots and whether such actions should be coordinated.
Garante prohibited AI chatbot company Replika from using the personal data of Italian users in February, citing risks to minors and emotionally fragile people.
