ChatGPT Becomes a "Search Engine": What the New EU Classification Means for Your Business

On August 31, 2026, the European Commission officially classified ChatGPT as a "Very Large Online Search Engine" (VLOSE) under the Digital Services Act. Alongside Reddit and Roblox, which were designated as "Very Large Online Platforms" (VLOP), OpenAI's AI chatbot now falls into a category previously reserved for just two services: Google Search and Bing. According to press reports, ChatGPT is thus the first AI chatbot to be classified as a search engine under the DSA – a first that goes far beyond a legal footnote.

For businesses using ChatGPT in their daily work, this initially sounds like an abstract regulatory announcement from Brussels. But it isn't quite. The classification says something fundamental about how ChatGPT now works – and it raises a question that is strategically relevant for SMBs: how visible is your business, really, when ChatGPT answers as a search engine? In this article, we break down what was actually decided, what obligations arise for OpenAI, what this means for your business – and what it doesn't.

What the EU Commission Has Specifically Decided

The Digital Services Act distinguishes between two top-tier oversight categories: "Very Large Online Platforms" (VLOP) and "Very Large Online Search Engines" (VLOSE). Both categories apply as soon as a service exceeds the statutory threshold of 45 million average monthly active users in the EU – roughly 10 percent of the EU population. Once a service crosses that threshold, it automatically falls under the strictest oversight tier the DSA knows, regardless of whether it's a social platform, a marketplace model, or an AI system with search functionality.

Press reports, including from Gizmodo and Euronews, put ChatGPT's search function's monthly EU user count at around 159 million over a six-month period through March 2026 – well above the threshold. According to these reports, Reddit comes in at around 57.2 million, and Roblox at around 48 million EU users. All three services have clearly exceeded the 45-million mark, which made the classification nearly unavoidable from a formal standpoint once the Commission had reviewed the user figures accordingly.

With ChatGPT, Reddit, and Roblox, the total number of services under DSA top-tier oversight rises to 28, according to reporting – comprising the 17 VLOPs originally listed in 2023 plus services added since, such as Bing and Wikipedia. What matters here is the distinction between the three newly added services: Reddit and Roblox were classified as platforms, while ChatGPT was explicitly classified as a search engine. That's not a formality but a functional categorization with its own logic – and this is exactly where it gets interesting for businesses.

Why ChatGPT Now Counts as a Search Engine

The reason for the VLOSE classification lies in how ChatGPT operates today. When the model answers a query, it increasingly draws on current internet content – via live web search and browsing functionality. From the EU Commission's perspective, ChatGPT thus behaves functionally like a classic search engine: it takes a user query, searches the open web, and delivers an answer based on current sources rather than relying exclusively on static training knowledge. This behavior is exactly what distinguishes today's version of ChatGPT from a pure language model that only answers from a fixed knowledge base.

OpenAI itself has publicly confirmed the classification, using a notably precise formulation. Speaking to the trade publication PYMNTS, the company stated:

"ChatGPT search operates as a search service under the DSA, and we are preparing to meet the additional compliance requirements that come with this Very Large Online Search Engine designation."

This statement is notable because OpenAI isn't referring to ChatGPT as a whole, but specifically to the search functionality that meets the DSA's search-engine definition. For businesses, this means: the part of ChatGPT your customers, your competitors, and prospective leads use to research today is now subject to the same regulatory standards as Google and Bing. That's a departure from earlier discussions about ChatGPT, which mostly revolved around content, training data, or copyright – here, for the first time, the focus is explicitly on its role as a research tool.

What Obligations Now Apply to OpenAI – and What That Means for You

The VLOSE designation brings a range of additional compliance obligations for OpenAI, going well beyond what applies to smaller providers. Under the DSA, these include in particular annual systemic risk assessments covering, among other things, the spread of illegal content, protection of minors, impacts on mental and physical health, fundamental rights, electoral processes, and public safety. Additionally, there are mandatory independent audits, which OpenAI must commission at its own expense, as well as data access for vetted external researchers, so that academia and civil society can independently examine how the service works.

Two further obligations are especially relevant for businesses: disclosure requirements on how ranking and recommendation systems work – in other words, the logic by which ChatGPT favors certain sources, answers, or content – and a non-personalized usage option, allowing users to opt out of profile-based ranking. Both obligations create more transparency around a mechanism that has so far largely been a black box.

Important for context: these obligations apply exclusively to OpenAI as the service provider – not to the businesses that use ChatGPT in their daily operations. If you use ChatGPT for research, customer communication, or content creation, this designation does not make you a regulated VLOSE operator yourself, and you don't need to fulfill any DSA reporting obligations of your own. Failure to comply with these obligations exposes OpenAI to fines of up to 6 percent of global annual revenue – a framework that generally applies to all VLOP and VLOSE violations of the DSA, not specifically to the ChatGPT case.

Reporting on the implementation deadline varies slightly: the DSA generally provides for a four-month implementation period after designation for the additional obligations, which would put it, calculated from August 31, 2026, at the end of December 2026. Individual press reports, however, cite "end of November 2026" or "January 2027" as the target date. A single, consistently documented exact deadline cannot currently be derived from the available sources – businesses tracking this development should therefore assume roughly "late 2026 to early 2027" rather than relying on a single date.

The Real Strategic Question: Visibility in AI Answers

The obvious first reaction from many business owners is: does this even affect me? The short answer is: barely directly, but quite a bit indirectly. The DSA regulates the provider, not the usage. It changes nothing about the obligations that already apply to your own AI usage anyway – under GDPR, the EU AI Act, or general competition law. If you use ChatGPT compliantly today, you'll continue to do so after August 31, 2026, with no change to your own obligations.

The indirect level is more interesting. More transparency obligations for OpenAI mean, over the medium term, more traceability of how and why ChatGPT names particular businesses, products, or sources in its answers. The disclosure requirements on ranking and recommendation logic, the mandatory audits, and the annual risk reports create – at least in theory – more insight into a mechanism that has so far remained largely hidden. That's relevant for any business that depends on being mentioned at all in AI-generated answers, whether it's a trade business, a law firm, a consultancy, or a manufacturer.

Because that's really the crux of it: when the EU Commission treats ChatGPT functionally like a search engine, it regulatorily confirms what many in marketing have long observed. For a growing share of your prospective customers, ChatGPT is already the first port of call for research – ahead of Google, ahead of your own website, ahead of the industry directory. The question is no longer whether this is happening, but how well your business is prepared for it when a prospect asks ChatGPT: "Which provider in my area is best at X?"

Classic search engine optimization has been a fixed part of the SMB marketing mix for decades: keywords, backlinks, technical SEO, local listings. The VLOSE classification of ChatGPT reinforces a trend that's already underway – the need to optimize your own web presence for discoverability by AI systems as well. The industry has coined a term for this: Generative Engine Optimization, or GEO. GEO follows partly different rules than classic SEO: it's less about keyword density and more about structured data, verifiable authority, mentions in trustworthy sources, and reviews and media coverage that a language model can cite as evidence of facts. A business with a consistent, up-to-date presence in trade articles, directories, and review platforms has a better chance of actually being named in a ChatGPT answer than one relying solely on classic keyword ranking.

With the new transparency obligations for OpenAI, this area could become significantly more concrete in the coming months. If OpenAI has to disclose the logic by which content is prioritized in answers, that will eventually create more clarity about which signals actually matter – similar to how Google updates have repeatedly reshaped the SEO industry in the past. Businesses that engage with this development early gain an edge over those that only react once traffic from AI answers has already noticeably dropped or risen.

One further, so far less-discussed aspect: should OpenAI introduce advertising formats within ChatGPT in the future, additional transparency obligations would apply as a VLOSE operator – such as a publicly accessible ad archive, as the DSA requires for large platforms. Whether and in what form ChatGPT currently runs any advertising at all is not conclusively confirmed, and corresponding claims remain speculative for now. Still, businesses looking to advertise within AI answers down the line would do well to keep an eye on this development – additional documentation and labeling requirements for advertisers are generally standard for VLOSE services.

What You Should Do Right Now

Rather than ignoring the classification or overreacting to it, it's worth taking a pragmatic look at three areas of action. First: check how your business currently appears – or doesn't appear – in ChatGPT's answers. Ask typical customer questions about your industry, your offering, and your competitors, and observe which sources are cited, which businesses are mentioned, and on what grounds. That gives you a first, easy-to-gather impression of where you stand today.

Second: keep an eye on the rollout of OpenAI's new obligations, especially the disclosures on ranking logic, as they're published. They can provide valuable clues about which signals actually matter for your own GEO strategy, rather than leaving you to guess. The annual risk reports and audit results that OpenAI will have to publish going forward should also gradually create more transparency about how the service works.

Third: avoid drawing a sense of urgency from this news that objectively isn't there. This is a regulation of the provider, not a new obligation or liability for your business. The DSA changes nothing about existing requirements under GDPR, the EU AI Act, or competition law that already apply to your own AI usage – much like the requirements that have applied to SMBs since August 2026 under the EU AI Act, which we broke down in a dedicated article. There, too, the same principle applies: the regulation primarily targets providers of AI systems, but businesses using AI tools should actively monitor developments rather than ignoring or overestimating them.

The strategically smarter response, then, is not alarmism, but a sober stocktaking: where does my business stand in terms of AI visibility, which content and sources already contribute to that visibility today, and what follows from that in the coming months? Whoever answers these questions now is better positioned than competitors who probably just skimmed past the news.

Conclusion: A Sign of Maturity, Not an Alarm

The classification of ChatGPT as a Very Large Online Search Engine is, first and foremost, a regulatory sign of maturity. It shows that the EU Commission no longer treats ChatGPT as an experimental tool, but as an established part of the digital infrastructure, subject to the same oversight and transparency standards as Google and Bing. Mandatory audits, risk assessments, and disclosure requirements tend to increase the reliability and auditability of ChatGPT as a research tool – an aspect that can build more long-term trust in the service's governance for businesses using ChatGPT for customer communication, internal research, or content creation.

For SMBs, little changes legally right away on August 31, 2026. Strategically, though, now is exactly the time to look closely: the confirmation that ChatGPT functionally works like a search engine is an invitation to take your own visibility in AI answers seriously – before your competitors do. Whoever understands today how ChatGPT responds to queries about their own industry has a real head start once this type of research takes hold among more customer segments.