EU AI Act Article 50

Article 50 of the EU AI Act establishes transparency obligations that apply to providers and deployers of AI systems generating synthetic audio, image, video, or text content. The rules become enforceable on 2 August 2026. Under Article 50, providers must ensure AI-generated outputs are marked in a machine-readable format detectable as artificially generated or manipulated. Deployers of deepfake systems must disclose that content has been artificially generated. Publishers of AI-generated text on matters of public interest must disclose the AI origin of the text. The European Commission published a second draft Code of Practice on AI content labelling and transparency in March 2026, with stakeholder feedback closing 30 March 2026.

Article 50 is distinct from the parent regulation (EU AI Act, in force since August 2024) in that it specifically targets the synthetic content pipeline rather than AI systems classified by risk tier. For broadcast media companies using AI-generated commentary, synthetic anchors, AI-assisted voiceovers, or automated news text, Article 50 imposes a compliance deadline concurrent with the peak summer news cycle. France Télévisions, which already uses Moments Lab AI tooling in its production workflows, faces a concrete obligation to label AI-generated content visible to French viewers from that date .

The SMART STORIES consortium's SOM architecture is directly relevant: if the Story Object Model carries a provenance field indicating AI-generated components, it could function as the machine-readable disclosure mechanism Article 50 requires, embedding compliance metadata at the story-object level rather than requiring post-production labelling on each output asset.