In a press release, the GDPD said it is investigating ChatGPT after banning the popular OpenAI chatbot.
Italy has become the first Western country to officially ban the AI chatbot ChatGPT over privacy concerns. The European country is seeking to launch an investigation into alleged violations of its data privacy rules. The investigation will also determine whether the OpenAI chatbot complies with the General Data Protection Regulation.
On Friday, the Italian data protection authority requested the suspension of ChatGPT following a platform violation on March 20. According to the Data Protection Authority (GDP), the data of one user was exposed to another user through an AI chatbot.
While accusing ChatGPT of leaking conversations and payment information of at least one user, the GDPD also raised another pressing question. The Italian data regulator pointed out that insufficient information was provided about the data compiled by OpenAI. In a press release, the security regulator said:
“The privacy guarantor observes that users and all interested parties of the data collected by OpenAI lack information, but most importantly, lack legal grounds to justify the collection and massive storage of personal data in order to ‘train’ the underlying algorithmic platform’s operations.”
Furthermore, the GDPD noted that the information presented by ChatGPT does not always reflect real data. The agency concluded that chatbots can be inaccurate when processing personal data.
Italian ChatGPT ban comes as data regulator claims chatbot lacks filters to verify user age
The GDPD also mentioned that ChatGPT may have violated its own data protection rules. The Italian data regulator explained that while the chatbot had an age restriction, it lacked filters to verify the user’s age. This operational flaw meant that people under the age of 13 could have been exposed to age-inappropriate answers.
After Italy banned ChatGPT, GDP said OpenAI had 20 days to fix the issues raised.
According to the watchdog: "OpenAI has appointed a representative in the European Economic Area and must notify within 20 days the measures taken to implement the guarantor's request, which carries a maximum fine of €20 million or 4% of its global annual turnover."
The ban is not the first time Italy’s data regulator has taken action against AI chatbots. In February, Italy’s data regulator banned chatbot app Replika from accessing personal data, citing risks to emotionally vulnerable people and minors.
Although ChatGPT had around 100 million users worldwide in its first two months, it is unclear how many users there are in Italy. An hour after the Italian ban was announced, some publications still claimed to be able to access the AI service from the country.
Despite its early success and explosive popularity, ChatGPT has also come under fire from other parts of the world. For example, the Center for Artificial Intelligence and Digital Policy (CAIDP) today filed a complaint against the chatbot. CAIDP described ChatGPT as a "biased" and "deceptive" platform that could harm public privacy.