EU classifies ChatGPT as search engine, first AI chatbot under DSA

The European Commission ended any ambiguity about how it sees generative AI's role in the digital public square. On Monday, it classified ChatGPT as a search engine and designated it a Very Large Online Search Engine under the Digital Services Act, the bloc's landmark content-moderation rulebook. That makes OpenAI's chatbot the first AI assistant ever placed in this enforcement tier, a decision that lands on a service used by hundreds of millions of people to answer questions, summarize documents, and generate creative work. No other AI chatbot has yet been forced into this category, and the implications for how OpenAI designs, documents, and monitors ChatGPT in Europe are immediate and structural. The Commission paired the move with simultaneous designations of Reddit and Roblox as Very Large Online Platforms, showing the DSA's net is being cast wider than the usual social media names. What was once a regulatory gray area for generative AI is now clearly marked: if a service helps people find information, it can be treated as a search engine, regardless of whether it crawls the web or draws on a trained model. That conceptual leap is what makes this announcement a watershed for the industry. The DSA is the EU's flagship legislation for policing online content, and its strictest duties apply only to platforms and search engines that reach a certain scale in Europe. Until now, those duties fell on the likes of TikTok, X, and Google Search. ChatGPT had been watched closely by regulators, but it did not carry the formal label. Monday's decision changes that. As a Very Large Online Search Engine, ChatGPT now carries the same systemic responsibilities as Google Search. That means OpenAI must conduct annual risk assessments, publish transparency reports, give vetted researchers access to data, and take measures to mitigate risks like the spread of election misinformation or the amplification of harmful content. Regulators can also request information directly and levy fines if they find systemic shortcomings, and enforcement can escalate quickly if the Commission concludes ChatGPT is contributing to societal harms. The designation also creates a precedent.