The European Commission has formally designated ChatGPT as a Very Large Online Search Engine, or VLOSE, under the Digital Services Act.

The designation was announced on August 31. Reddit and Roblox were designated on the same day as Very Large Online Platforms.

ChatGPT is the unusual one because the Commission placed it in the search-engine category.

That designation activates the DSA's more demanding framework for services operating at systemic scale.

The legal threshold starts at 45 million monthly EU users

The DSA's size threshold corresponds to roughly 10% of the European Union's population.

The Commission says ChatGPT declared that it reaches at least 45 million average monthly users in the EU and therefore meets the condition for designation.

The relevant figure is not an external web-traffic estimate. It comes from user information reported for regulatory purposes.

European reporting has indicated that ChatGPT's disclosed reach is in practice substantially higher than the minimum threshold.

Legally, however, the important point is simply that the threshold has been crossed and the Commission has issued the designation.

Why is a chatbot being regulated as a search engine?

The label initially sounds awkward for a product that became famous as a conversational assistant.

The DSA is not concerned with OpenAI's preferred marketing category.

Its definition is functional: a service that allows users to submit internet search queries and returns information related to those requests can fall within the online-search-engine framework.

Web search has become a significant capability inside ChatGPT.

The Commission has therefore concluded that the service falls within the VLOSE regime.

That does not mean every ChatGPT feature is suddenly legally identical to Google Search or Bing.

It means its qualifying search functionality is sufficient to bring the service into this part of the DSA.

OpenAI now has four months

Designation does not make every additional requirement effective overnight.

The DSA gives newly designated VLOPs and VLOSEs four months after notification to comply with the enhanced obligations.

For an August 31 designation, that places the compliance deadline around the end of 2026.

OpenAI was already subject to general DSA requirements applicable to its European service.

It already maintains an EU authority contact point and publishes information covering content reporting, moderation decisions, appeals and DSA transparency.

The August designation adds the systemic-risk layer.

The question is no longer only whether one piece of content is illegal

VLOSE obligations require a provider to assess risks created by the service and its algorithmic systems as a whole.

The Commission explicitly lists dissemination of illegal content, negative effects on minors, users' physical and mental wellbeing, fundamental rights, electoral processes and public security.

OpenAI will have to examine how its systems may contribute to those risks and document measures designed to mitigate them.

For a conventional search engine, systemic analysis can focus heavily on ranking and recommendation.

ChatGPT creates a different problem because the system may synthesize the underlying information and generate a single conversational answer instead of simply presenting links.

How that distinction is treated in risk assessments may become one of the most consequential parts of the designation.

Minors and mental wellbeing are explicitly inside the scope

The Commission's announcement goes beyond disinformation and clearly illegal material.

It also identifies potential negative effects on minors and on physical and mental wellbeing.

That can push regulation beyond traditional content moderation.

Relevant questions may involve product design, generated recommendations, interactions with vulnerable users and safeguards intended to prevent the service from reinforcing harmful behavior.

The designation does not itself prescribe a specific answer to every one of those problems.

It requires OpenAI to assess them and demonstrate mitigation.

Election integrity becomes a formal systemic-risk category too

Electoral processes are explicitly included in the DSA's systemic-risk framework.

That principle already applies to large social platforms and conventional search engines.

Applying it to a generative assistant creates another dimension.

A user can ask directly about a candidate, political claim or public institution and receive a synthesized response rather than a simple set of sources.

Potential risk therefore includes not only whether false information is available but how the system summarizes and prioritizes it.

The Commission did not accuse ChatGPT of causing a specific election-related incident in the August 31 designation.

It is bringing the service inside the preventive framework designed to address those risks before they become systemic.

The European Commission becomes the direct supervisor of the enhanced rules

Designation also changes the enforcement structure.

The European Commission directly supervises the enhanced DSA obligations applying to VLOPs and VLOSEs.

It can request information, investigate compliance and open formal proceedings where it suspects breaches.

For serious violations, the DSA allows fines of up to 6% of a provider's worldwide annual turnover.

That figure is a statutory maximum, not a fine that has been imposed on OpenAI.

It illustrates the enforcement power attached to the category ChatGPT has now entered.

The DSA and AI Act regulate different layers

ChatGPT's designation arrives only weeks after another major EU AI regulatory milestone.

From August 2, 2026, additional AI Act provisions entered application or enforcement, including transparency requirements affecting some AI interactions and artificially generated or altered content.

The same product can therefore fall under both laws without the two rules doing the same job.

The AI Act regulates AI systems and models according to their characteristics, uses and risks.

The DSA examines ChatGPT as a public-facing digital service, including its scale, algorithmic systems and broader societal effects.

A particular issue may consequently fall under one framework or both depending on what causes the risk.

Europe is beginning to regulate AI assistants as information infrastructure

That may be the more important consequence of the Commission's decision.

Early generative-AI regulation focused heavily on how models are trained, what data they use and what individual answers they produce.

VLOSE designation adds another perspective.

ChatGPT is now used widely enough to be treated as one of the interfaces through which a significant share of the public searches for and receives information.

That is why audience scale matters.

An incorrect answer generated inside an experimental prototype and the same error reproduced by a service used by tens of millions of Europeans do not create the same systemic exposure.

The DSA does not yet provide the answer to every regulatory question created by that transition.

It has now decided that ChatGPT is large enough to be required to measure them.