ROME/STOCKHOLM (Reuters) – Italy’s data protection agency Galante has said ChatGPT producer OpenAI has taken steps to address concerns that led to Italy’s ban on chatbots last week. It said it will be presented on Thursday.
Microsoft Corp-backed (MSFT.O) OpenAI has taken ChatGPT offline in Italy after Garante temporarily restricted it last week and has launched an investigation into alleged privacy rule violations.
Last week, the agency accused OpenAI of failing to check the age of ChatGPT users and of having “no legal basis to justify the large-scale collection and storage of personal data.”
On Thursday, it said it would not put the brakes on AI development, but reiterated the importance of respecting rules aimed at protecting the personal data of Italian and European citizens.
In a videoconference attended by CEO Sam Altman late Wednesday, Garante said OpenAI pledged to be more transparent about how it handles user data and how it verifies the age of its users.
The company said it will send Garante a letter on Thursday regarding steps it will take to respond to the request.
The Data Authority said it appreciates the proposal by OpenAI. A source familiar with the matter said it would likely take several days to evaluate the contents of the letter.
San Francisco-based OpenAI did not respond to a request for comment on the agency’s statement.
On Thursday, the company published a blog post titled “Working on AI Safety,” saying it was working on “a nuanced policy for behavior that represents a real risk to people.”
“We do not use the data to sell our services, advertise, or build profiles of people,” it said. For example, ChatGPT will improve with more training on talking to people.
“Some of our training data includes personal information available on the public internet, but we want our model to learn about the world rather than individuals.”
The company said it would remove personal information from its datasets when possible, fine-tune its model to reject user prompts for such information, and respond to individual requests to remove data from its systems. rice field.
Italy’s ban has attracted the attention of other European privacy regulators who are considering whether tougher measures against chatbots are necessary and whether such measures should be coordinated.
In February, Garante banned AI chatbot company Replika from using Italian users’ personal data, citing risks to minors and emotionally vulnerable people.
Reported by Supantha Mukherjee from Stockholm, Elvira Pollina from Milan and Martin Coulter from London. Written by Keith Weir. Edited by Alvise Armellini, Jason Neely, Alexander Smith, Christina Fincher
Our standards: Thomson Reuters Trust Principles.