ChatGPT designated as search engine under EU Digital Services Act rules

The European Commission has officially classified ChatGPT as a Very Large Online Search Engine under the Digital Services Act. The move also designates Reddit and Roblox as Very Large Online Platforms, marking the first time an AI chatbot has been brought under this specific regulatory regime.

The classification triggers automatically for services exceeding 45 million monthly users in the European Union. ChatGPT reported 159 million monthly active users in the six months leading to March 2026, far surpassing Reddit’s 57.2 million and Roblox’s roughly 48 million.

Treating a prose-generating chatbot as a search engine represents a novel regulatory interpretation. The Digital Services Act imposes distinct obligations based on service type, meaning OpenAI must now apply search-related compliance frameworks to a system that generates answers rather than merely ranking web pages.

These companies must now submit to annual systemic risk assessments covering illegal content, fundamental rights, electoral processes and public security. They are also required to allow independent audits and grant data access to regulators and vetted researchers.

“These new designations mean that ChatGPT, Reddit and Roblox will now be held to a higher standard of scrutiny and accountability,” said Henna Virkkunen, the Commission’s executive vice-president for tech sovereignty, security and democracy. This heightened oversight arrives as data access remains a contentious issue, particularly when researchers seek information that companies consider commercially sensitive.

The designations apply to named entities rather than consumer-facing brands, specifically listing OpenAI Ireland Limited, Reddit Netherlands B.V. and Roblox Corporation. These companies have until the end of November to comply, with serious violations risking fines of up to 6 per cent of global annual turnover.

Divergent compliance challenges

Each platform faces distinct regulatory hurdles under the new framework. Roblox must formally assess and mitigate risks to children and teenagers, who make up a large portion of its audience, with these safety measures subject to independent auditing.

OpenAI must translate broad concerns about AI influence into measurable risks surrounding electoral processes and public security. Meanwhile, Reddit’s designation complicates its recent commercial strategy of licensing its vast public discussion data to AI developers, as researchers will now demand greater access to that same information.

Enforcement responsibilities will be shared between the Commission and national regulators in Ireland and the Netherlands. None of the three companies have publicly challenged the designations, though appealing would require them to dispute the user numbers they themselves reported.

This move signals that European technology enforcement is expanding well beyond traditional social networks. As the Commission simultaneously pursues action against Meta over scam advertising and prepares to force Google to open Android to rival AI assistants, the November risk assessment deadlines will test whether legacy platform rules can effectively govern generative artificial intelligence.