EU classifies ChatGPT as a 'Very Large Online Search Engine'

[Joker16]

The European Commission classified ChatGPT as a Very Large Online Search Engine under the EU's Digital Services Act on August 31st, putting OpenAI's chatbot under the same strict regulatory tier as established search engines like Google and Bing. The designation follows OpenAI reporting approximately 159.1 million average monthly users in the EU, more than three times the 45 million user threshold that triggers this level of oversight. Reddit and Roblox were separately designated as Very Large Online Platforms under the same rules

The Commission based the search engine classification specifically on ChatGPT's interactive web search features, the functionality letting users ask questions and receive answers synthesized from live internet sources, arguing that capability makes it functionally equivalent to a traditional search engine for regulatory purposes. OpenAI now faces compliance obligations including systemic risk assessments covering illegal content, protection of minors, and impacts on users' mental health, independent audits, and data access requirements for researchers, with compliance required by January 2027 or fines of up to 6 percent of global annual turnover

European Commission Executive Vice President Henna Virkkunen said the new designations mean ChatGPT, Reddit and Roblox will now be held to a higher standard of scrutiny given their large impact on citizens and society. The designation is a legal classification triggering new obligations rather than a finding that ChatGPT has broken any law. Curious what people think about regulators applying search engine rules to an AI chatbot, does that framework actually fit how people use conversational AI, or does it force an awkward regulatory square peg into a round hole

EU classifies ChatGPT as 'Very Large Online Search Engine' under DSAwww.business-standard.com

[Sinead]

6 percent of global annual turnover is a genuinely serious penalty ceiling, that's not a fine companies can just treat as a cost of doing business, it's large enough to actually force real compliance investment

[QuoteMiner22]

Curious how OpenAI's mental health and minor protection risk assessments will actually work in practice for a conversational AI product, those obligations were clearly written with traditional search and social platforms in mind first

[MutedAgain57]

This being a legal classification rather than a finding of wrongdoing is worth remembering, it's easy to read headlines like this as EU punishing ChatGPT when it's really just triggering a new oversight tier based on scale alone

[Louise5]

159.1 million monthly users obliterating the 45 million threshold rather than just barely crossing it makes this designation feel less like a close judgment call and more like an inevitable regulatory outcome

[ObserverEffect68]

Basing the classification specifically on the web search feature rather than ChatGPT's core conversational function is a clever bit of regulatory reasoning, letting the Commission apply an existing legal category rather than needing to write an entirely new one from scratch

[PlanckLimit65]

January 2027 gives OpenAI a real runway to build out compliance infrastructure, though four months from notification to deadline is still a tight timeline for the kind of systemic risk assessments this designation requires