The European Commission has designated ChatGPT a "Very Large Online Search Engine" under the Digital Services Act, the first time a standalone AI chatbot has been placed in a regulatory category previously reserved for Google Search and Bing. The classification, made on August 31, triggers a new set of oversight duties for OpenAI's flagship product across the bloc.
The designation rests on scale and function rather than how OpenAI markets the product. ChatGPT reported roughly 159.1 million average monthly users in the EU, more than three times the 45 million threshold that triggers the DSA's toughest tier of scrutiny. The Commission classified the chatbot as a "hybrid service" that qualifies as a search engine because it can respond to prompts by actively searching the web — a capability-based test that regulators could apply just as easily to Google's Gemini, Anthropic's Claude or Perplexity as those products scale their own EU user bases.
Crucially, the label is not a finding that OpenAI has broken any rules. It is a legal classification that starts the clock on compliance obligations, and enforcement would require a separate proceeding if regulators later find the company falling short.
What OpenAI now has to do
Under the "Very Large" designation, OpenAI must conduct annual systemic risk assessments covering illegal content, harms to minors, mental and physical wellbeing, fundamental rights and effects on civic discourse and elections. It also has to submit to independent external audits, publish transparency reports, and give EU regulators and vetted outside researchers conditional access to data for studying those risks. Companies typically have around four months from formal notification to fall in line, putting OpenAI's deadline in the window between late 2026 and early 2027, depending on which Commission timeline is used.
The financial exposure for falling short is significant: fines can reach 6% of a company's global annual turnover. Regulators have already set a baseline for how seriously they intend to enforce the DSA's largest-platform tier — X was fined €120 million in December 2025 under the same framework. The European Commission holds exclusive supervisory authority over Very Large designees, while Ireland's Coimisiún na Meán, as OpenAI's lead national regulator, will help coordinate researcher access requests.
For OpenAI, the designation adds a second major EU compliance track alongside the AI Act, which regulates the company's models directly rather than the product built on top of them. Together, the two regimes mean ChatGPT is now subject to more formal European oversight than any other AI chatbot on the market.
What happens next will be watched closely well beyond OpenAI's offices. Because the Commission's reasoning turns on a feature — live web search — rather than a company or a specific product category, it effectively gives regulators a template they can reuse. Any AI assistant that answers questions by searching the live web and clears the user threshold could face the same classification, turning what looks today like an OpenAI-specific ruling into the opening move of a broader campaign to bring conversational AI under the DSA's largest-platform rules.
