In June 2026, the European Commission published the final Code of Practice on the Transparency of AI-Generated Content, developed by the AI Office through a multi-stakeholder process. The Code is a voluntary instrument designed to support compliance with Article 50 of Regulation (EU) 2024/1689, the EU AI Act, which requires providers of AI systems to ensure their outputs are marked in a machine-readable and interoperable format as artificially generated or manipulated. The Article 50 obligations become applicable on 2 August 2026.
The Code distinguishes two groups of obligations. For providers, Article 50(2) requires that audio, image, video, and text outputs of AI systems be technically marked in a machine-readable and interoperable format detectable as artificially generated. For deployers, Article 50(4) requires that AI systems used to generate or manipulate images, video, or audio constituting deepfakes disclose this clearly to exposed persons, and that AI-generated text published for the purpose of informing the public on matters of public interest be labelled as AI-generated. The Code sets specific technical and procedural commitments for signatories in each category.
AI system providers placing products on the EU market and deployers using AI for public-information purposes must assess their obligations under Article 50 ahead of the 2 August 2026 application date. Signatories to the Code benefit from a streamlined enforcement approach: future AI Office monitoring will focus on adherence to Code commitments, delivering greater legal certainty and reduced administrative burden across EU member states. News publishers, social media platforms, audiovisual media services, and AI-generated content providers are the principal addressees.
The AI Omnibus, formally adopted by the EU Council on 29 June 2026, introduces a grace period until 2 December 2026 for AI systems placed on the market before 2 August 2026. Providers of legacy systems therefore have until December 2026 to implement compliant marking, but new systems placed on the market from 2 August 2026 must be compliant from that date. The Code does not substitute for national authority enforcement, which remains governed by the AI Act's supervisory chapter.
Licentium advises AI providers and deployers on EU AI Act transparency compliance. If your organisation is assessing its Article 50 obligations or evaluating whether to sign the Code of Practice, we and our partner network can assist. Work we undertake includes AI Act transparency obligation mapping, machine-readable marking implementation guidance, deepfake disclosure compliance, and regulatory strategy for AI providers and deployers in media, financial services, and consumer-facing sectors.