European Commission publishes AI transparency code second draft

The European Commission has released the second draft of its Code of Practice on Transparency for AI-Generated Content, following feedback from industry, academia, and civil society. This revised version aims to reduce the compliance burden by providing greater flexibility and technical clarity for signatories. Key updates include promoting open standards for content marking and proposing the development of a uniform EU icon for labelling.

The code is divided into two sections. Section 1 focuses on generative AI providers and introduces a streamlined, two-layered approach to marking content through watermarking and secure metadata. Section 2 targets the deployers of AI systems, specifically addressing the labelling of deepfakes and text relating to matters of public interest. This section has been simplified by removing the distinction between AI-generated and AI-assisted content, instead focusing on practical requirements for icons and disclaimers. Specific regimes for creative, satirical, and fictional works have been further defined to align with existing editorial practices. 

The Commission will collect further feedback until 30 March 2026, with the final code expected in June, ahead of transparency rules established under Article 50 of the Artificial Intelligence Act (AI Act), set to become applicable on 2 August 2026. 

Training Announcement: The BCS Foundation Certificate in AI examines the challenges and risks associated with AI projects, such as those related to privacy, transparency and potential biases in algorithms that could lead to unintended consequences. Explore the role of data, effective risk management strategies, compliance requirements, and ongoing governance of the AI lifecycle and become a certified AI Governance professional. Find out more.