EU designates ChatGPT, Reddit and Roblox very large services
Developed over time first seen 2 months ago
The European Commission has designated OpenAI's ChatGPT, Reddit and Roblox as "very large" online services under the EU's Digital Services Act, placing them under the bloc's strictest online safety obligations. ChatGPT was classified as a "Very Large Online Search Engine" while Reddit and Roblox were designated "Very Large Online Platforms", after each surpassed 45 million average monthly users in the EU. The designation brings closer scrutiny of how the platforms' algorithms handle illegal content, protect minors and safeguard electoral processes, reflecting Brussels' push to extend established digital regulation to fast-growing AI and social platforms.
The three firms now face four months to comply, including sharing data with regulators, undergoing annual independent audits and granting researchers access to platform data in certain cases. A Reddit spokesperson said the company was "committed to complying" and had been preparing to meet the requirements. The additions bring the total number of services designated as very large platforms or search engines under the DSA to 28, following WhatsApp's designation earlier this year and those already applied to Facebook, Instagram and Google's search engine.
- EU designates ChatGPT, Reddit and Roblox as "very large" online services under the DSA
- Each platform passed 45 million average monthly EU users, triggering stricter rules
- Firms get four months to comply with audits, data-sharing and researcher access rules
New here? Start with this
This EU rule change concerns the Digital Services Act (DSA), a law that sets out how large online platforms must handle harmful or illegal content and protect users, particularly children. Once a platform crosses 45 million monthly users in the EU, it is classed as a "very large online platform" and must follow the DSA's strictest requirements, backed by fines of up to 6% of global revenue for non-compliance.
OpenAI's ChatGPT, the discussion site Reddit and the gaming platform Roblox have all now been placed in this category by the European Commission, alongside long-standing entrants such as major social media and shopping sites. This is notable because it is one of the first times an AI chatbot has been brought under these rules, showing how EU regulators are extending existing internet laws to cover fast-growing generative AI tools rather than treating them separately.
The designation matters because it signals how the EU intends to police AI-driven services going forward, at a time when it is also rolling out separate AI-specific legislation and facing criticism from the United States over its treatment of American tech firms. The companies involved now have a matter of months to show they meet the new obligations.
Both sides, in good faith
The strongest fair case each way — we don't pick a winner.
The case for
Supporters of the EU's approach argue that any platform reaching tens of millions of users, including a chatbot, carries the kind of societal influence that warrants the strictest safeguards, particularly around illegal content and the protection of children. They see applying the same threshold-based rules to ChatGPT as were applied to established social networks as principled consistency rather than targeting, especially since the Commission has also fined a Chinese firm under the same regime. From this view, regulation maturing in step with a technology's reach is a reasonable, precautionary way to build public trust in fast-evolving AI services.
The case against
Critics contend that imposing the DSA's toughest tier of obligations on a general-purpose AI assistant risks conflating fundamentally different products, since a chatbot's risks and moderation needs differ from those of a social network or gaming platform, and heavy compliance costs could slow innovation or entrench only the largest incumbents able to absorb them. Some also see the pattern of enforcement, following scrutiny of X's Grok and probes into Google, as lending credibility to Washington's complaint that Brussels is disproportionately targeting American technology firms. They would prefer rules tailored to how generative AI actually poses risk, rather than applying an existing framework built for older platform types by default.
Full account
OpenAI's ChatGPT, Reddit and the gaming platform Roblox have been brought under the strictest tier of the European Union's online safety regime, the Digital Services Act (DSA), the European Commission announced on Monday. All three services have been designated after crossing the threshold of at least 45 million average monthly users in the bloc, a scale that automatically triggers heightened regulatory obligations. ChatGPT has been classified as a "Very Large Online Search Engine", while Reddit and Roblox have been designated "Very Large Online Platforms" — categories that, together, now cover 28 services under the law, alongside established names such as Facebook, Instagram and, since earlier this year, WhatsApp.
The new status brings a raft of additional duties. The three platforms must act more assertively to remove illegal content, shield minors from harmful material and protect the privacy and security of younger users, and assess how their algorithms and recommender systems could affect matters such as electoral processes. They will also face closer oversight from Brussels, including obligations to share data with the Commission and national authorities, submit to annual audits by independent reviewers, and give vetted researchers access to platform data. Henna Virkkunen, the Commission's executive vice-president for tech sovereignty, security and democracy, said the designations meant the three services would "be held to a higher standard of scrutiny and accountability", reflecting their scale of impact on EU citizens and society. Companies that fail to comply risk fines of up to 6% of their global annual revenue.
The firms have been given until the end of the year to bring their operations into line — a period reported as roughly four months from the announcement. OpenAI said it was preparing to meet the new requirements, while Roblox said it was "proud to be the first gaming platform to reach this milestone in the EU" and pointed to safety as central to its operations, adding that it looked forward to continued engagement with the Commission. Reddit had not issued a public response by the time the news broke, though reports differ slightly on whether all three companies had yet replied to requests for comment.
The move extends the DSA's reach further into generative artificial intelligence, following a separate investigation the Commission has opened into Elon Musk's X over its Grok chatbot. It also comes as Brussels intensifies wider scrutiny of major technology firms: the bloc is beginning to enforce its AI Act, has opened a competition probe examining whether Google's AI Overviews breach EU rules, and fined the Chinese marketplace AliExpress €550 million in July over failures to curb illegal goods. That backdrop has fuelled tension with Washington, which has accused Brussels of unfairly targeting American companies and of infringing free-speech principles associated with the MAGA movement — a charge the Commission rejects, insisting its digital rules apply uniformly regardless of where a company is headquartered.
Where outlets differ
The Financial Times describes all three services simply as newly classified 'very large online platforms', while the other report is more precise in noting that ChatGPT was specifically designated a 'Very Large Online Search Engine', a distinct DSA category from the 'Very Large Online Platform' label applied to Reddit and Roblox.
The FT situates the news within a much wider geopolitical and regulatory narrative — the Grok investigation, the EU AI Act's enforcement, the Google AI Overviews antitrust probe, the AliExpress fine, and transatlantic tension with Washington over the DSA — whereas the other outlet focuses narrowly on the procedural mechanics of the designation itself.
The other report gives more granular detail on the compliance machinery (independent audits, researcher data access, the running total of 28 designated services, and the prior addition of WhatsApp), detail largely absent from the FT account.
On timing, the FT states firms must comply 'by the end of December', while the other outlet frames this as a four-month compliance window; the two are consistent but expressed differently.
The FT includes direct company reactions from OpenAI and Roblox and notes Reddit did not immediately respond, while the other outlet states it had reached out to all three companies and was awaiting comment, without direct quotes.
More coverage
Read the full article at the source →
Originally published by Ars Technica as “ChatGPT and Reddit now face EU’s toughest online safety rules”.