EU publishes AI Act transparency guidelines
Guidance for Article 50, which takes effect August 2, tells providers how to label AI-generated content and disclose AI interactions.
The European Commission published guidelines on July 20 for Article 50 of the AI Act, which applies from August 2 and requires providers to design generative and interactive AI systems so users know they’re interacting with AI, and to add machine-readable marks to AI-generated or altered content, including deepfakes. The guidance is aimed at reducing deception risk as the deadline approaches.
Update, August 2: The Commission confirmed enforcement of these rules started today. More than 180 organizations have signed the EU’s Code of Practice on AI-generated content as a recognized way to demonstrate compliance, and the Commission has published dedicated complaint and whistleblower tools, including a separate channel for downstream providers building on general-purpose AI models. The obligations described above are no longer a pending deadline, they’re active now.