OpenAI must now mitigate risks related to ChatGPT’s impact on minors, user mental health, and the spread of illegal content within the European Union. The European Commission has classified the service as a Very Large Online Search Engine under the Digital Services Act, placing it under stricter oversight than before. This regulatory shift joins Reddit and Roblox in the list of Very Large Online Platforms facing similar obligations. The new rules prohibit these services from targeting advertisements toward minors or using sensitive personal data such as sexual orientation, religion, ethnicity, or political beliefs for marketing purposes. OpenAI will be held accountable for ensuring compliance with these provisions, which aim to reduce the potential harm of generative AI systems on vulnerable populations. The designation forces the company to implement more rigorous content moderation and transparency measures regarding how the model interacts with children and processes user information. This move signals that EU regulators view large language models as high-risk environments requiring direct intervention rather than relying solely on voluntary industry standards.
• The Digital Services Act applies specifically to services with over 45 million monthly active users in the EU.
• OpenAI must publish detailed reports on how it mitigates risks to minors and mental health.
• Ad targeting based on sensitive personal data is now strictly prohibited for all designated platforms.
