What's Happening?
The European Commission has officially designated OpenAI's ChatGPT as a Very Large Online Search Engine (VLOSE) under the Digital Services Act (DSA). This classification applies to online services with at least 45 million average monthly users in the EU
that retrieve and synthesize web content in response to user queries. ChatGPT's reported 159.1 million monthly average users in the EU significantly exceed this threshold. As a VLOSE, ChatGPT will face stringent compliance obligations, including annual systemic risk assessments, independent audits, researcher data access, and maintaining a public repository of advertisements. This designation sets a precedent for other AI search products like Google Gemini and Perplexity, which are also likely to meet the user threshold or are already subject to similar national regulations, such as those in Germany.
Why It's Important?
This designation is a landmark decision that significantly impacts the regulatory landscape for AI services, particularly those operating in the U.S. and globally with a European user base. It establishes a legal template for how AI-powered search and conversational tools will be governed, emphasizing accountability and transparency. For U.S. tech companies, this means that their AI products, regardless of their primary classification (chatbot or search engine), will be subject to rigorous EU oversight if they reach a certain scale. The compliance requirements, such as systemic risk assessments covering illegal content, harms to minors, and threats to electoral processes, will necessitate substantial investment in engineering, legal, and organizational infrastructure. This could influence product development and operational strategies for AI companies aiming for international markets, potentially leading to a global standard for AI regulation.
What's Next?
OpenAI must now build the necessary infrastructure to meet the DSA's compliance obligations by January 2027. This includes conducting annual systemic risk assessments, undergoing independent audits, providing data access to vetted researchers, and establishing a public ad repository. The technical challenges are considerable, especially for algorithmic transparency, as generative AI models do not operate with traditional ranking parameters. OpenAI will need to work with Ireland's Coimisiún na Meán, its national supervisor, to interpret these requirements for LLM-based services. The EU's enforcement track record, with significant fines issued for DSA violations, indicates that non-compliance carries substantial financial penalties. Other AI search products, including Google Gemini and Perplexity, are now on notice to prepare for similar designations as they scale, with Germany's national media law already applying to AI search outputs regardless of user thresholds.
Beyond the Headlines
The EU's VLOSE designation for ChatGPT highlights a critical debate about the nature of AI and its regulatory classification. By treating ChatGPT as a search engine due to its web-retrieval function, the EU is signaling that the functional capabilities of AI, rather than its self-description, will determine its regulatory treatment. This 'capability-based' logic has profound implications for the development and deployment of AI globally, pushing companies to consider regulatory compliance from the outset of product design. The dual regulation under both the DSA and the EU AI Act creates a complex compliance environment, requiring companies to address both the underlying AI model and the service built upon it. This comprehensive approach aims to mitigate systemic risks associated with large-scale AI deployment, fostering a more responsible and transparent AI ecosystem, but also posing significant challenges for innovation and operational agility for U.S. tech companies.
