EU places ChatGPT under strict Digital Services Act rules
The European Union has designated ChatGPT as a very large online search engine, forcing OpenAI to meet strict safety and transparency rules to protect millions of European users.

The European Commission has officially designated OpenAI's ChatGPT as a Very Large Online Search Engine under the Digital Services Act. Alongside the popular AI chatbot, the EU also classified social media platform Reddit and gaming ecosystem Roblox as Very Large Online Platforms. This regulatory shift places all three services under the highest tier of European digital oversight.
Under the Digital Services Act framework, digital services receive this designation once their user base crosses a threshold of at least 45 million average monthly users within the European Union. Because of this massive reach, OpenAI, Reddit, and Roblox must fully align their operations with the stringent requirements of the law. The European Commission has set a deadline of late December 2026 for these companies to achieve complete compliance.
The new classification forces OpenAI to actively mitigate systemic risks associated with ChatGPT. Specifically, the company must address the chatbot's potential impact on minors, user mental health, and the dissemination of illegal content. Additionally, the law prohibits these platforms from targeting advertisements to minors or leveraging sensitive personal data, such as sexual orientation, religion, ethnicity, or political beliefs. It also mandates greater transparency regarding how their internal recommendation algorithms function.
For AI practitioners and developers building on or competing with OpenAI, this designation signals a major shift toward mandatory transparency. Henna Virkkunen, the EU's executive vice-president for tech sovereignty, security, and democracy, noted that the designated platforms will now be held to a higher standard of scrutiny and accountability due to their societal footprint. Engineers and product managers working with large-scale models in Europe must now prepare for rigorous auditing of safety guardrails, algorithmic transparency, and data privacy protections.
This is our own summary of reporting by The Verge AI



