Italy Bans ChatGPT
Italy has banned OpenAI’s generative AI chatbot ChatGPT. The Italian Data Protection Authority ordered OpenAI to halt any operations in the country and stop processing data from Italian users while it looks into possible violations of Europe’s privacy regulations.
ChatGPT Privacy Problems
The Italian regulatory agency, also called Garante, pointed to the recently uncovered and resolved ChatGPT vulnerability that allowed some users to see the titles of conversations others had with the chatbot. OpenAI took ChatGPT down to fix the problem and later posted a blog post explaining what happened and how it was fixed. Nonetheless, Garante cited the event as an example of its concerns, tied to broader legal issues about how ChatGPT collects and uses information for training and fine-tuning the underlying large language model. The agency has begun looking into whether OpenAI has violated Europe’s General Data Protection Regulations. GDPR encompasses data storage and usage rules for companies.
Garante shared a statement saying that there “appears to be no legal basis underpinning the massive collection and processing of personal data in order to ‘train’ the algorithms on which the platform relies.”
The agency also cited ChatGPT’s lack of age barriers as another potential example of OpenAI flouting GDPR. Notably, Google’s new Bard generative AI chatbot is only available to those who assert they are over 18 and accept legal liability. OpenAI has only missed brief statements in reply so far, insisting that it does comply with all relevant privacy laws. OpenAI has 20 days to produce solutions to Garante’s claims or could face a 20 million Euro fine, nearly $22 million.