# EU Classifies ChatGPT as a Search Engine — What Developers Must Know > Source: > Published: 2026-09-02 09:15:28+00:00 On August 31, 2026, the European Commission designated ChatGPT as a Very Large Online Search Engine (VLOSE) under the Digital Services Act — the first time a generative AI product has been regulated as a search engine. The trigger: 159.1 million average monthly EU users, more than triple the 45-million threshold. OpenAI now has until the end of November 2026 to comply with the EU’s strictest digital safety obligations, or face fines of up to 6% of global annual turnover — a number that, at OpenAI’s current revenue trajectory, could exceed $1.2 billion. ## Why “Search Engine” Matters More Than You Think The EU did not classify ChatGPT as a search engine because OpenAI called it one. It classified it based on what it *does*: the Commission determined ChatGPT qualifies because it “can engage with and respond to users’ prompts and queries, including by searching the web.” Product category is irrelevant. Capability is everything. That distinction is the most important thing to understand here. The EU has established a capability-based template — not a product-category one. Any AI service with live web search and 45 million or more EU monthly users is a VLOSE candidate under this framework. Gemini has 950 million global monthly active users and almost certainly clears the EU threshold already. Perplexity is sitting right at the global 45-million boundary. Claude’s EU user count is undisclosed but growing fast. The [regulatory cascade](https://digital-strategy.ec.europa.eu/en/news/commission-designates-chatgpt-reddit-roblox-under-digital-services-act) is not a future possibility — it is already in motion. ## What VLOSE Compliance Actually Requires Being designated a VLOSE is not just a label. It comes with the EU’s heaviest compliance tier. OpenAI must now: - Conduct **annual systemic risk assessments** covering illegal content, protection of minors, user wellbeing, fundamental rights, electoral integrity, and public security - Submit to **independent annual audits** of DSA compliance - Grant **data access to vetted researchers** studying systemic risks - Maintain a **searchable ad repository** with full details: content, topic, targeting parameters, and the legal entity paying for each ad - Offer users **at least one non-profiling recommender option**— a feed that doesn’t rank based on their behavioral profile - Publish **twice-yearly transparency reports** with aggregated, anonymized metrics - Maintain a **crisis response mechanism** for public security emergencies The deadline for the heaviest obligations is the end of November 2026. Four months. That is not a lot of time for a company the size of OpenAI to instrument compliance infrastructure across a product used by 159 million EU users. ## The Enforcement Precedent Is Real Some EU tech regulation functions as a soft warning. DSA enforcement does not. On December 5, 2025, the Commission issued its [first-ever DSA fine](https://ec.europa.eu/commission/presscorner/detail/en/ip_25_2934): €120 million against X/Twitter for three specific violations — selling verified checkmarks without identity verification, maintaining a sparse and non-compliant ad repository, and blocking mandated researcher access in its terms of service. The EU has accumulated €870 million in total DSA fines to date. The €120 million X fine is the floor, not the ceiling. For OpenAI, the math is sobering. At an estimated $20 billion-plus annual revenue run rate, a 6% fine works out to over $1.2 billion. The Commission has demonstrated it follows through. ## What This Means If You’re Building on ChatGPT Here’s the part most coverage has gotten wrong: if you are an API consumer — a developer building on top of the ChatGPT API — the VLOSE obligations sit with OpenAI, not with you. You are not directly subject to DSA compliance requirements by virtue of using their API. That said, the downstream effects are real and worth planning for now: **Expect UI/UX changes in EU-facing ChatGPT.** Feed controls, moderation transparency labels, and non-profiling options are all coming. If your product wraps or embeds ChatGPT’s interface in any form, the UX will shift.**Watch for API terms updates in Q4 2026.** OpenAI will almost certainly update EU-facing API terms as it instruments compliance. Read the changelogs.**Researcher access APIs create opportunities.** The DSA mandates data access for vetted researchers studying systemic risks. If you build safety or audit tooling, this is a new market opening. The more forward-looking implication: treat the 45-million-user threshold as a design constraint for any EU-ambition AI product you build. Retrofitting audit logging, moderation transparency, and ranking explainability onto a deployed system at scale is far more expensive than building those capabilities in from the start. [OpenAI’s own DSA documentation](https://help.openai.com/en/articles/8959649-eu-digital-services-act-dsa) offers a starting point for understanding how they are framing compliance obligations. ## Who’s Next in the Regulatory Cascade The EU Commission’s capability-based classification template does not stop at ChatGPT. Based on current user trajectories, Gemini is the most obvious next candidate — Google has not disclosed EU-specific monthly active user numbers, but at 950 million global MAU, it almost certainly clears the threshold. Perplexity is hovering at the global boundary. Claude’s EU numbers are growing. Meanwhile, the DSA VLOSE designation is separate from — and parallel to — the EU AI Act, which has its own compliance track for high-risk AI systems. Major AI providers are now navigating dual regulatory frameworks simultaneously. That complexity will surface in pricing, API terms, and product roadmaps through 2027. For a closer look at how the DSA’s VLOSE obligations are defined, [DSA Article 33](https://dsa-library.com/article/33/) lays out the designation criteria in full. The bottom line: the EU just made AI search a regulated utility. Build accordingly.