EU eyeing tighter OpenAI rules
European regulators are considering whether ChatGPT should be classed as a ‘large online platform’ under the Digital Services Act, which would impose stronger obligations on risk management, transparency and oversight. If that classification sticks, operators would face heavier compliance duties that could change how AI products are built and marketed in Europe. The shift from debate to potential enforcement is already prompting firms to rethink investment and operational plans in the region. (reuters.com)
European officials are weighing whether ChatGPT belongs in the same regulatory bucket as the biggest social platforms, a move that would pull OpenAI into the toughest part of the European Union’s Digital Services Act. Reuters reported on April 10, 2026 that the question inside Brussels is whether ChatGPT should count as a “large online platform.” (reuters.com) That label is not just a name. Under the Digital Services Act, services with at least 45 million monthly users in the European Union face the strictest oversight, including annual risk reviews, outside audits, and direct supervision by the European Commission. (digital-strategy.ec.europa.eu, eur-lex.europa.eu) The law was written with places like marketplaces, app stores, and social networks in mind. Europe is now testing whether a chatbot that answers questions, writes code, and generates images should be treated like a platform that distributes information at mass scale. (digital-strategy.ec.europa.eu, eur-lex.europa.eu) The reason is reach. The European Commission says the threshold for a “very large online platform” is 10% of the European Union population, which it currently describes as 45 million users, and the whole system is built around the idea that very large services can create society-wide risks that smaller ones cannot. (digital-strategy.ec.europa.eu, digital-strategy.ec.europa.eu) If ChatGPT lands in that category, OpenAI would not just file paperwork. It would have to assess and reduce systemic risks tied to illegal content, effects on fundamental rights, and manipulation of civic processes, then submit to independent audits on whether those controls actually work. (eur-lex.europa.eu, digital-strategy.ec.europa.eu, eur-lex.europa.eu) Europe has already started writing generative artificial intelligence into its Digital Services Act playbook. The Commission’s own guidelines say very large platforms should adopt election-specific measures related to generative artificial intelligence, which shows regulators are no longer treating chatbots as a side issue. (digital-strategy.ec.europa.eu) This would sit on top of Europe’s separate Artificial Intelligence Act, not replace it. The Digital Services Act governs how big digital services operate in public, while the Artificial Intelligence Act governs how high-risk artificial intelligence systems are built and used. (digital-strategy.ec.europa.eu, eur-lex.europa.eu) That overlap helps explain why companies are nervous. OpenAI has spent the past two years building a deeper European footprint with operations in Dublin, a Brussels office aimed at policy engagement, and a Munich office in Germany, even as the rulebook around it keeps getting tighter. (openai.com, brusselstimes.com, openai.com) The fight is really about what ChatGPT is. If regulators decide it is not just a tool you open one-on-one but a mass distribution system that shapes what millions of Europeans read, ask, and believe, then the European Union’s biggest platform rules start to look like a match. (reuters.com, digital-strategy.ec.europa.eu) And once that door opens for OpenAI, it does not stay limited to one company. A decision that a chatbot can be a “large online platform” would give Brussels a template for judging rivals like Google’s Gemini, Anthropic’s Claude, and Meta’s assistant as their user numbers climb in Europe. (reuters.com, digital-strategy.ec.europa.eu)
