Italian ban on ChatGPT may be resolved with remedies proposed by the maker

Italy's data protection agency Garante said that ChatGPT producer OpenAI plans to present measures on Thursday to address concerns that led to the chatbot's ban last week.

‍

As part of a probe into a suspected breach of privacy rules, OpenAI, backed by Microsoft Corp (MSFT.O), has temporarily taken ChatGPT offline in Italy after Garante, the Italian government agency, temporarily restricted it last week.

‍

It was alleged last week by the agency that OpenAI failed to verify the age of ChatGPT users and that the "massive collection and storage of personal data done by the company had no legal basis in any case".

‍

A spokesperson for the Italian government told the media on Thursday that it has no intention of slowing down the development of artificial intelligence, but reiterated the importance of respecting rules designed to protect the personal data of Italian and European citizens.

‍

During a video conference called by OpenAI's CEO Sam Altman late on Wednesday, Garante said that OpenAI has pledged to become more transparent about the way in which it handles user data and verifies the user's age.

‍

Apparently, the company said it is sending Garante a document regarding measures to be taken in response to the company's requests by Thursday at the latest.

‍

OpenAI has indicated that it will evaluate the proposals that have been presented by the data authority. Several days will probably need to pass for the contents of the letter to be assessed, according to a source familiar with the matter.

‍

Despite a request for comment about the agency's statement, OpenAI, which is based in San Francisco, did not respond to the request for comment.

‍

As the company announced on Thursday, in a blog post titled "Our approach to AI safety", the company stated that it was working on developing "nuanced policies that will represent a real risk to people."

‍

The company asserts that it does not use personal data for selling its services, advertising, or building profiles of its customers. "We use data to improve our models so they can be more useful to people. ChatGPT, for instance, is improved by further training based on the conversations people have with the model."

‍

"Even though some of our training data includes personal information that is available on the Internet, we want our models to learn about the world as a whole, not about the people in it."

‍

As a result, the company said it had removed personally identifiable information from its datasets when possible, fine-tuned its models so that they would reject the user's requests for such information, and would respond to individual requests to have their data deleted.

‍

A ban imposed by Italy on chatbots has attracted the attention of other European privacy regulators who are investigating whether harsher measures should be taken against chatbots and whether such measures should be coordinated.

‍

The Italian data protection agency Garante banned AI chatbot company Replika from using the personal data of Italians, citing the possibility of exposing minors and people with fragile emotional states to harm.