OpenAI and its flagship product, ChatGPT, are facing scrutiny from Italy’s Data Protection Authority, also known as the Garante. The Garante accused OpenAI of breaching the European Union’s General Data Protection Regulation (GDPR), primarily terms dictating the handling of young users’ data and content moderation.
GDPR, a stringent privacy law in the EU, allows for hefty fines – up to 4% of a company’s global revenue – for non-compliance. This situation poses significant challenges for OpenAI, which has already experienced a temporary ban of ChatGPT in Italy last year, marking the first such action in Europe. Although the app was reinstated after OpenAI addressed the initial privacy concerns, it has still been under close observation from an EU task force since April.
Key issues raised by the Garante include the collection of user data for training the chatbot and the potential exposure of underage users to inappropriate content. OpenAI’s platform requires users to be at least 13 years old, with those under 18 needing parental or guardian consent. Despite these measures, the Italian firm remains concerned about the effectiveness of these policies in safeguarding young users.
In response, OpenAI has asserted its commitment to aligning with GDPR and other privacy laws. The company has emphasized its efforts to minimize the use of personal data in training its systems and to reject requests for private or sensitive information.
This development in Italy is part of a broader trend of increasing legal and regulatory scrutiny of AI technologies globally. OpenAI, backed by tech giant Microsoft, has not only faced challenges in Europe but also in the United States. Antitrust watchdogs in both regions are examining the relationship between Microsoft and OpenAI for potential competition concerns. Additionally, OpenAI has been embroiled in a copyright infringement lawsuit filed by the New York Times for allegedly using its articles to train ChatGPT without proper authorization.
Europe’s proactive stance to legislating AI products stands in stark contrast to the U.S.; as the market develops there are few frameworks to follow, and large-language-models will continue to do whatever possible to stay above their competitors.
