ChatGPT under the microscope: The EU classifies AI as a search engine
The European Commission marks a regulatory milestone by including ChatGPT in the Digital Services Act, redefining the future of generative AI.
September 3, 2026 · 3 min read
TL;DR: The EU has classified ChatGPT as a very large search engine, integrating it into the strict rules of the Digital Services Act. This step forces OpenAI to raise its standards for transparency, safety, and auditing on European soil.
A paradigm shift in digital regulation
The European Commission's recent decision to classify ChatGPT as a Very Large Online Search Engine (VLOSE) under the Digital Services Act (DSA) represents the most ambitious regulatory move since the creation of the GDPR. By equating the AI's search functionality with that of giants like Google or Bing, Brussels establishes a historic precedent: the technical distinction between a language model (LLM) and an information search engine is, for legal purposes, obsolete. This change is not merely semantic; it is a recognition that user behavior has migrated from hyperlink navigation to conversational querying, shifting the power of web indexing toward probabilistic inference.
Why is ChatGPT now a search engine?
The DSA, initially designed for platforms like Meta or X, requires entities with more than 45 million monthly active users in the EU to implement strict systemic risk management protocols. According to Henna Virkkunen, Executive Vice-President of the European Commission, the integration of real-time search capabilities has transformed ChatGPT's architecture. By allowing the model to access the web index to answer queries, OpenAI has ceased to be a static language processing engine to become a central node for global information access. Unlike a traditional search engine that returns a list of URLs, ChatGPT processes, synthesizes, and occasionally alters the presentation of information, which, according to the Commission, amplifies risks such as mass disinformation and algorithmic bias. This designation places OpenAI in a regulatory category shared with players like Reddit and Roblox, albeit under distinct technical scrutiny due to its generative nature.
Impact on the startup ecosystem and large companies
The VLOSE designation implies an unprecedented operational burden. For OpenAI, this means the obligation to subject its models to independent external audits, a process that was opaque until now. The challenge is both technical and ethical: how can a company ensure that a model that 'hallucinates' by design complies with the EU's requirements for 'informational accuracy'? For SaaS startups integrating OpenAI APIs, this creates legal uncertainty. If the underlying model must comply with the DSA, companies implementing it could be forced to perform due diligence on the traceability of training data far deeper than what the previous framework dictated. Unlike the era of the classic search engine (where responsibility partially fell on the linked site), under the DSA, the engine that synthesizes the response is directly responsible for the integrity of the content. This could raise the barrier to entry for new competitors, consolidating the position of incumbents who possess sufficient capital to fund massive legal compliance teams.
AI regulation is not a brake on innovation, but a necessary infrastructure to ensure safety in the age of generative algorithms.
Consequences for the future of work and AI
We are witnessing the end of the 'Wild West era' in LLM development. This regulatory convergence suggests that the future of search will not be a list of links, but a curated response, which forces companies to rethink their traditional SEO strategy. For knowledge workers, this implies greater transparency regarding the origin of data, reducing the risk of hidden biases in responses. However, market speculation suggests that this measure could slow down the deployment of advanced features in the EU compared to markets like the US. It is highly likely that other leading models, such as Anthropic's Claude or Google's Gemini, will follow similar regulatory trajectories. While Google is already under scrutiny for its search engine, its generative AI division will now have to navigate a dual framework: the DSA for its search facet and the future AI Act for its generative capacity. This regulatory overlap is uncharted territory that will define the commercial viability of the next generation of digital tools in the European bloc.