What happened
The European Commission's guidelines on the transparency obligations under Article 50 of the AI Act were published in the Official Journal (OJ C/2026/6005) on 12 November 2026 (in-window). Adopted under Art. 96(1)(d) of Regulation (EU) 2024/1689, the non-binding guidelines clarify obligations for human-AI interaction, machine-readable marking of synthetic content, emotion recognition, biometric categorisation, and disclosure of deepfakes/AI-generated public-interest content.
Why it matters
This is a major implementation milestone for the EU AI Act, giving explicit, EU-authoritative operational guidance on a set of obligatunctions that became applicable in August 2026. Even though it is a transparency/disclosure instrument rather than a security control catalogue, it is part of the same risk-management and evidence surface that AI-security frameworks must satisfy, and it establishes machine-readable marking/tagging requirements that affect AI system design and content pipelines.
Action needed
Map the Article 50 transparency obligations against your AI system controls and documentation; confirm your chatbot/AI-assistant disclosure, synthetic-content marking, and deepfake disclosure mechanisms satisfy the guidance; align evidence collection with ISO/IEC 42001 / NIST AI RMF to avoid duplicative assurance.