EU Commission Takes Aim at Illegal AI Content with New Guidelines

The European Commission on Thursday published draft guidelines for so-called trusted flaggers under the Digital Services Act (DSA). These specialized organizations help online platforms identify illegal content faster. The move could also have major implications for AI platforms, which increasingly face AI-generated fraud, copyright violations, and other illegal content.

According to the Commission, online platforms must prioritize reports from trusted flaggers. This includes child sexual abuse material, online fraud, intellectual property rights, and other forms of prohibited content. Platforms remain responsible for the final assessment of whether the reported content is actually illegal.

The new guidelines arrive as generative AI plays a growing role in creating and spreading online content. AI models can produce large volumes of text, images, videos, and audio in seconds. That drives innovation—but also makes it harder to catch illegal content in time.

For providers of AI chatbots, AI search engines, and generative image and video models, the trusted flagger system could become pivotal. Reports from recognized organizations can trigger faster reviews of content distributed or amplified by AI systems.

The European Commission stresses that trusted flaggers do not have the power to remove content themselves. Their reports do, however, move to the front of platforms’ moderation queues.

The development shows how different European digital laws are becoming increasingly intertwined. While the AI Act focuses on the development, deployment, and risks of AI systems, the DSA targets the spread of content via digital services.

In practice, those lines are blurring. AI systems generate content, while platforms are responsible for distributing it. As a result, companies building AI face not only requirements on model safety and transparency, but also obligations around content moderation and risk management.

For big tech companies, that shifts the focus from pure technical performance to broader responsibilities around safety, oversight, and compliance with European regulation.

The draft guidelines also set additional requirements for organizations designated as trusted flaggers. They must operate independently, act objectively, and publish annual transparency reports on their activities.

The guidelines also introduce safeguards against abuse of the system. Regulators can suspend or revoke trusted flagger status if organizations no longer meet the criteria.

According to the European Commission, more than seventy trusted flaggers have now been designated across the European Union. These organizations cover a wide range of expertise, from financial fraud and online harassment to combating child sexual abuse material.

For AI companies, this is another clear signal that Europe is tightening control over digital content. As generative AI becomes more central to online ecosystems, the pressure grows to detect and remove illegal content faster.

Together, the DSA and the AI Act create a regulatory framework that increasingly scrutinizes both the creation and distribution of AI content. For AI service providers, effective content moderation is quickly becoming a strategic necessity rather than a nice-to-have.

Stakeholders can submit feedback on the draft guidelines until 26 June 2026. The European Commission expects to adopt the final version in the second half of 2026.