At a glance: Article 50 mandates AI transparency disclosure in four scenarios: direct AI-human interaction, synthetic content, emotion and biometric recognition, and deepfakes and AI-generated text on matters of public interest.
Article 50 of the EU AI Act obligates providers and users of AI systems from 2 August 2026 to implement transparency requirements — regardless of whether these systems are classified as high-risk. The regulation effectively affects every organisation that uses generative AI for content creation.
Article 50 of the EU AI Act establishes transparency obligations in four concrete situations: when AI interacts directly with humans, when AI generates synthetic content, when AI is used for emotion recognition or biometric categorisation, and when AI creates deepfakes or text published on matters of public interest. These obligations apply to all AI systems in these four situations — not only to systems classified as high-risk. The obligations come into force on 2 August 2026.
For providers, this results in concrete implementation obligations: chatbots, virtual assistants and other interactive systems must be designed so that users recognise they are communicating with AI. Generative AI systems must mark their outputs in machine-readable format and make them identifiable as artificially generated — a standardised EU marking scheme is being developed. Users of emotion or biometric recognition must inform affected persons. Those creating deepfakes must disclose their artificial nature. Those publishing AI-generated text on matters of public interest must mark this, unless the text has been reviewed by humans and editorially verified.
A particular feature of Article 50 is its broad applicability regardless of high-risk status: an organisation without high-risk AI can still have substantial obligations — for example, through customer-facing chatbots, AI tools for news generation, or deepfake systems. Compliance data shows that transparency obligations under Article 50 are the second most common compliance challenge and affect approximately 33 per cent of organisations surveyed. For many, Article 50 will thus become the primary regulatory challenge.
The European Commission has published draft guidelines on the interpretation and application of Article 50. A code of conduct on AI-generated content is being developed to provide practical solutions for marking and labelling. An important exemption exists when it is obvious that AI is involved — from the perspective of an appropriately informed, attentive and prudent person. The guidelines recommend a two-step approach here: first assessment of the target audience, then verification of surface perception.
Source: ainews-dev.lumi-systems.io · Published 23 May 2026
Lumi AI News — AI-assisted curation pursuant to Article 50 EU AI Act. Paraphrase and classification by Lumi News Pipeline v1.5.2.