European Commission Publishes Code of Practice on Labelling AI-Generated Content Ahead of August AI Act Obligations

On June 10, 2026, the European Commission published the final version of its Code of Practice on marking and labelling of AI-generated content¹. The code is a voluntary framework setting out practical steps to help providers and deployers of generative AI systems comply with the EU AI Act’s transparency obligations, which apply from August 2, 2026³.

With less than two months until the obligations take effect, the document serves as a concrete preparation guide for companies offering generative AI in the EU, as well as those creating and distributing AI-generated content there.

The Transparency Obligations Under Article 50 of the AI Act

The code addresses the transparency obligations that Article 50 of the AI Act imposes on providers and deployers of generative AI systems, targeting the risks of AI-driven deception and manipulation².

The obligations differ by role. Providers are required to mark AI-generated content in a machine-readable way and enable its detection, while deployers must disclose deepfakes and AI-generated or manipulated text on matters of public interest². Specifically, deepfakes must be clearly labelled, and AI-generated or manipulated text published on matters of public interest also requires clear labelling¹. Users must additionally be informed when they are interacting with an interactive AI system, such as a chatbot¹.

For AI-generated text, an exemption applies where a human exercises editorial responsibility over the content².

How the Code Is Structured and Why Companies Might Sign

The code of practice consists of two sections. Section 1 is aimed at providers and sets out rules for marking and detection of AI-generated and manipulated content. Section 2 is aimed at deployers and covers rules for labelling deepfakes and AI-generated and manipulated text².

While the code is voluntary, signatories will collaborate in Signatory Taskforces and, in return, gain predictability, legal certainty, and trust across all EU member states². The Commission has also published a page explaining how to sign the code, along with “EU icons” that can be used to label AI-generated content¹.

Drafted Through More Than Half a Year of Multi-Stakeholder Consultation

The code was drafted by independent experts in a multi-stakeholder process facilitated by the Commission’s AI Office. Participants included providers of generative AI systems, developers of marking and detection techniques, associations of deployers, and civil society organizations².

The process began with a consultation in September 2025, followed by a kick-off plenary on November 5, 2025, leading to the publication of the final version on June 10, 2026².

What Companies Should Check Now

The August 2 start of the AI Act’s transparency obligations is not just an issue for companies based in the EU. Japanese and other non-EU companies that offer generative AI services to the EU, or distribute AI-created content there, may also be affected. A practical first step is to determine whether your organization qualifies as a “provider” or a “deployer,” and whether the content you handle includes deepfakes or AI-generated text on matters of public interest, then check it against the code’s rules.

The full text of the code (PDF) and signing instructions are available on the European Commission’s policy page².