On Friday, Italy announced that it would temporarily block ChatGPT due to concerns regarding data privacy. This move makes Italy the first western nation to take action against the well-known AI chatbot.
The Italian Data Protection Authority stated that OpenAI, the US-based company behind ChatGPT, had no legal justification for the mass collection and storage of personal data to train the platform's algorithms. ChatGPT gained global attention for its ability to produce essays, songs, exams, and news articles from short prompts. However, concerns about the data sources and processing methods of ChatGPT and its competitors have persisted. Some education authorities and universities have prohibited the use of the chatbot, fearing that it may be used to cheat in exams or write essays. Recently, several experts and industry figures signed an open letter advocating for a halt in the development of powerful AI systems, citing their potential risks to society and humanity. OpenAI's release of GPT-4, a more advanced version of ChatGPT with less transparency regarding its data sources, was the catalyst for the letter.
OpenAI has responded to Italy's decision to block ChatGPT by disabling the service for users in Italy. An OpenAI spokesperson stated that the company is committed to protecting individuals' privacy and believes that it complies with privacy laws. The spokesperson added that OpenAI works to minimize personal data in training its AI systems, including ChatGPT. They also acknowledged the need for AI regulation and expressed willingness to work closely with Italian authorities to educate them on how their systems are built and used. The Italian Data Protection Authority has launched an investigation into OpenAI, citing a lack of legal basis for data collection and unclear data ownership. The authority also criticized the chatbot's wrong and "absolutely unsuitable answers," which it claimed exposed children to inappropriate content. Despite this, OpenAI says that Italian users have expressed their appreciation for ChatGPT and looks forward to making the service available again in the future.
The Italian Data Protection Authority cited a data breach on March 20 in which user conversations and payment information were compromised, an issue that OpenAI blamed on a bug. While securing user data and enforcing age limits were easy to address, the other two allegations were more challenging. Specifically, the watchdog criticized the use of personal data gathered without consent and not treated properly during ChatGPT's training. Nello Cristianini, an AI academic from Bath University, believes that these issues may not be resolved anytime soon. OpenAI has 20 days to respond to the investigation and may face a fine of up to €20 million ($21.7 million) or 4% of its annual revenue. Despite these concerns, OpenAI's successful development of ChatGPT has led to a lucrative deal with Microsoft, which integrates the technology into its Bing search engine and other programs. Other tech firms and venture capitalists have also rushed to invest in AI projects, with Google quickly launching its own chatbot.