EU eyes OpenAI under DSA
European regulators are assessing whether OpenAI should be treated as a “large online platform” under the Digital Services Act, which would layer platform duties on top of AI rules. If true, that means obligations like tougher content governance, transparency and systemic-risk reporting could apply to products such as ChatGPT after OpenAI reported user numbers above the threshold. (reuters.com)
Europe’s internet rulebook was built for giants like Google and Meta, and now Brussels is asking whether ChatGPT belongs in the same bucket after OpenAI reported user numbers above the legal line. On April 10, 2026, the European Commission said it was analyzing whether ChatGPT should count as a large online platform under the Digital Services Act. (reuters.com) That line is 45 million monthly users in the European Union. The Commission’s own guidance says services above that level can be designated as “very large online platforms” or “very large online search engines,” and then get the toughest set of duties in the law. (digital-strategy.ec.europa.eu) OpenAI’s own European Union disclosure now says ChatGPT search had about 120.4 million average monthly active recipients in the six months ending September 30, 2025. OpenAI publishes that figure under Article 24(2) of the Digital Services Act through OpenAI Ireland Limited. (help.openai.com) The fight is partly about what ChatGPT is in legal terms. If regulators treat ChatGPT search as an online search engine instead of just an artificial intelligence assistant, OpenAI could be pulled into a regime written for services that help people find and navigate information across the web. (help.openai.com) (digital-strategy.ec.europa.eu) The Digital Services Act is not the European Union’s artificial intelligence law. It is a platform law that covers things like illegal content reporting, terms of service, transparency around moderation, and risk controls for giant services that can shape what millions of people see. (digital-strategy.ec.europa.eu 1) (digital-strategy.ec.europa.eu 2) That matters because OpenAI is already inside the European Union’s artificial intelligence rulebook too. The Commission says obligations for providers of general-purpose artificial intelligence models started applying on August 2, 2025, with Commission enforcement powers starting on August 2, 2026. (digital-strategy.ec.europa.eu) So this is not one law replacing another. It is the possibility of two layers landing on the same company at once: one layer for the model underneath ChatGPT, and another layer for the consumer-facing service people use to search, read, and share information. (digital-strategy.ec.europa.eu 1) (digital-strategy.ec.europa.eu 2) If ChatGPT is formally designated, the clock gets short fast. The Commission says a service has four months after designation to comply with the special duties for very large platforms or search engines. (digital-strategy.ec.europa.eu) Those duties go beyond taking down bad posts. The Commission says very large services must assess systemic risks tied to illegal content, fundamental rights, public security, elections, gender-based violence, public health, minors, and mental and physical wellbeing, then put mitigation measures in place. (digital-strategy.ec.europa.eu) The practical question for OpenAI is what “mitigation” looks like inside a chatbot that can answer, summarize, and search in one box. The Commission’s examples for giant services include changing how the service is designed, changing recommendation systems, and adding internal compliance resources, which hints at product and policy changes rather than a simple paperwork exercise. (digital-strategy.ec.europa.eu) OpenAI has already built some of the plumbing Brussels expects. Its Digital Services Act page names a government contact point, routes official requests through the Kodex portal, and separates content-removal requests from user-data requests, which shows the company has been preparing for platform-style oversight before any formal designation. (help.openai.com) What Brussels decides next will say a lot about where Europe draws the line between an artificial intelligence model and an internet platform. If ChatGPT is treated as both, Europe will be telling every fast-growing artificial intelligence product that once it starts functioning like a search and distribution system, it can be regulated like one too. (reuters.com) (digital-strategy.ec.europa.eu)