LONDON (AP) 鈥 Italian regulators said they told OpenAI that its ChatGPT artificial intelligence chatbot has violated European Union鈥檚 stringent data privacy rules.
The country's data protection authority, known as Garante, that it notified San Francisco-based OpenAI of breaches of the EU rules, known as General Data Protection Regulation.
The watchdog started investigating ChatGPT last year, when it that can produce text, images and sound in response to users' questions.
Based on the results of its 鈥渇act-finding activity," the watchdog said it 鈥渃oncluded that the available evidence pointed to the existence of breaches of the provisions鈥 in the EU privacy rules.
OpenAI, which has 30 days to reply to the allegations, said it would work constructively with Italian regulators.
鈥淲e believe our practices align with GDPR and other privacy laws, and we take additional steps to protect people鈥檚 data and privacy," a company statement said. "We want our AI to learn about the world, not about private individuals. We actively work to reduce personal data in training our systems like ChatGPT, which also rejects requests for private or sensitive information about people.鈥
The company said last year that it that the Garante demanded to .
The watchdog had imposed the ban after finding that some users鈥 messages and payment information were exposed and because ChatGPT didn't have a system to verify users鈥 ages, allowing children to get answers from the AI tool that were inappropriate for their age.
It also questioned whether there was a legal basis for OpenAI to used to train ChatGPT鈥檚 algorithms and raised concerns that the system could sometimes generate false information about individuals.
The growing popularity of are also drawing increasing scrutiny from regulators on both sides of the Atlantic.
The U.S. Federal Trade Commission last week into the relationships between AI startups OpenAI and Anthropic and the tech giants that have bankrolled them 鈥 Amazon, Google and Microsoft. Competition regulators in the 27-nation EU and Britain, meanwhile, are also .
AI systems also face broader oversight in the EU, which is , the world's first comprehensive rulebook for artificial intelligence. The bloc's 27 member states are expected to endorse a version of the legislation Friday.
