AIToday
Large Language ModelsAI Regulation & PolicyTHE DECODERPublished: Sep 1, 2026, 01:00 JST

EU classifies ChatGPT as very large search engine

EU classifies ChatGPT as very large search engine

3 Key Points

  1. What happened

    The EU Commission classified ChatGPT as a very large search engine for the first time, announced Monday in Brussels. ChatGPT has at least 45 million monthly users in the EU. Reddit and Roblox were also reclassified as very large online platforms.

  2. Why it matters

    The classification triggers stricter oversight under the Digital Services Act. The three services have four months, until the end of December 2026, to meet extra obligations, including assessing risks like illegal content, dangers to minors, or election interference.

  3. What to watch

    The EU gains more audit powers, working with authorities in Ireland and the Netherlands. Legal experts debate whether data access could extend to training data or model weights, as law professor Natali Helberger noted that Article 40 access obligations might raise interesting questions.

Not sure about something? Ask the AI

Questions and answers are published on this page.

Summaries like this, in your inbox every morning.

Context & Analysis

The EU's move marks a significant step in applying the Digital Services Act to AI chatbots, recognizing that ChatGPT's web search and user query answering function constitutes a search engine. The threshold of 45 million monthly EU users triggered the classification, placing ChatGPT alongside 28 other platforms and search engines now under DSA oversight.

For OpenAI, the new rules introduce concrete compliance burdens, from ad transparency to researcher data access. The dispute over training data access shows the law's boundaries remain fuzzy; as Professor Helberger suggests, Article 40's access rights could potentially extend to model weights if necessary for risk assessment. The Commission's audit powers, exercised with Irish and Dutch authorities, will likely shape how these provisions are interpreted in practice.

FAQ
When do the new obligations take effect?
The services have four months, until the end of December 2026, to meet the extra obligations.
What does the classification require from ChatGPT?
ChatGPT must assess risks like illegal content, dangers to minors, or election interference. It also needs a public ad archive, researcher data access, transparency reports, and a complaints process.
Can the EU access ChatGPT's training data or model weights?
The classification doesn't explicitly allow direct model testing. Legal experts dispute whether Article 40 data access could include training data or weights if needed to identify systemic risk.

Also reported by The Verge AI

AI news that matters for your work, delivered every morning.

Pick your industry and the AI tools you use, and get news related to your work every day.

Free · 30 seconds with Google · unsubscribe anytimeWhat is AIToday? →

Ask AI

Ask AI anything about this article. The AI reads this article, earlier AIToday articles, and Wikipedia, and cites its sources. Q&As are published on this page for other readers too.

Questions and answers are published on this page.

Related Articles

Next articleDHH Releases Omarchy 4.0 with Built-in AI Agent