The European Commission has published a second draft of its voluntary Code of Practice on the marking and labelling of AI-generated content, a key step in helping companies comply with transparency requirements under the European Union’s Artificial Intelligence Act. This EU AI Act Code of Practice will be important for guiding future responsibilities and standards for organisations working with AI across the region.
Released March 5, the updated draft incorporates feedback from hundreds of stakeholders, including industry representatives, academic researchers, civil society groups and policymakers. The revisions follow consultations conducted through surveys, workshops and meetings held in January 2026 as part of the broader implementation process for the AI Act.
The code of practice is designed to assist both developers and deployers of generative AI systems in meeting obligations outlined in Article 50 of the law, which requires transparency when content is generated or manipulated by artificial intelligence.
The updated draft has been streamlined to reduce compliance burdens while offering greater flexibility for organizations that adopt the voluntary framework. It also encourages the use of open technical standards for identifying AI-generated content and proposes the creation of a standardized EU icon that could be used for labeling purposes across platforms and services.
The code is divided into two sections. The first section focuses on providers of generative AI systems and outlines technical approaches for marking AI-generated content. The revised draft introduces a two-layered system combining secured metadata and digital watermarking, along with optional measures such as fingerprinting, logging and verification protocols to help detect AI-generated material.
The second section addresses deployers of AI systems, including platforms and publishers, and outlines guidelines for labeling deepfakes and AI-generated text related to matters of public interest. The updated version removes earlier distinctions between AI-generated and AI-assisted content and instead focuses on practical labeling requirements, such as icon placement and disclosure formats.
The Commission will gather additional feedback on the draft until March 30, with a final version expected by early June. The AI Act’s transparency rules for AI-generated content will take effect across the EU on Aug. 2, 2026.
Need Help?
If you have questions or concerns about any global guidelines, regulations and laws, don’t hesitate to reach out to BABL AI. Their Audit Experts can offer valuable insight, and ensure you’re informed and compliant.
