Article 50 of 11344%
Article 50: Transparency obligations for providers and deployers of certain AI systems
Praxikon tracks Article 50 (Transparency obligations for providers and deployers of certain AI systems) under the EU AI Act, citing the source for every statement.
EU Official:
ActiveApplies since 2 Aug 2026
Title IV: Transparency ObligationsApplies from: 2 August 2026
Article 50 applies since 2 August 2026 and is the transparency article for AI interaction, chatbots, deepfakes, emotion recognition and AI-generated content. For European providers and deployers, the practical issue is the right timing, placement and wording of disclosures and labels. Systems already on the market before 2 August 2026 have until 2 December 2026 for machine-readable marking.
Official text
||
Source: EUR-Lex, Regulation (EU) 2024/1689. Text reproduced verbatim.
Download AI Act (PDF) →Official guidance on this article
13- Guidelines on transparency obligations for providers and deployers of certain AI systems under Article 50 of the AI ActExplains when the notification obligation for direct AI interaction applies, who precisely is provider or deployer, and what exceptions exist such as content that has only been subject to standard processing.PublishedGuidelinesEuropean Commission20 Jul 2026
- Code of Practice on Transparency of AI-generated ContentProvides the Commission-recognised voluntary route to demonstrate that the marking and labelling obligations for AI-generated content have been fulfilledPublishedCode of practiceEuropean Commission / AI Office10 Jun 2026
- EU icons for labelling AI-generated contentProvides a standardised visual form for the disclosure required by paragraph 4, so that people can see at a glance that content has been artificially generated or manipulatedPublishedTemplateEuropean Commission / AI Office6 Aug 2026
- Transparency obligations under Article 50 of the AI Act - Questions and AnswersExplains which notification obligation applies to direct interaction with an AI system, what exceptions there are, and who bears the obligation: the provider or the deployer.PublishedQ&AEuropean Commission, DG CONNECT24 Jul 2026
- Questions and answers on the Code of Practice for General-Purpose AIClarifies the distinction between the transparency rules for general-purpose AI models and the transparency obligations for AI systemsPublishedQ&AEuropean Commission, DG CONNECT20 Jul 2026
- EDPB Work Programme 2026-2027 (announcing joint guidelines on the interplay between the AI Act and the GDPR)Is to clarify the relationship between transparency obligations under Article 50 and information obligations towards data subjects under the GDPR, which the Commission already points forward to in its Article 50 guidelines.AnnouncedGuidelinesEDPB (European Data Protection Board)11 Feb 2026
- Guidance on generative AI and the GDPRIs relevant alongside the transparency obligation for generated content, because the same generative application must both be marked and requires a GDPR legal basisPublishedGuidelinesDutch Data Protection Authority (AP)13 Jul 2026
- Transparency requirements for AI: what does this mean for you?Translates the four transparency obligations into recognisable everyday situations and explains that the information must be available at the latest at the time of first interaction.PublishedQ&ADutch Data Protection Authority (AP)31 Jul 2026
- AI transparency requirements apply from 2 August: AP advises signing the code of practiceLinks each of the four obligations to the correct paragraph: paragraph 1 for direct interaction with an AI system, paragraph 2 for machine-readable marking of generated output, paragraph 3 for emotion recognition and biometric categorisation, paragraph 4 for visible labelling of deepfakes.PublishedNational guidanceDutch Data Protection Authority (AP), Algorithm Coordination Directorate9 Jul 2026
- Supervision of AI takes shape: key role for the AP and the RDIAnnounces that the AP receives supervisory authority for transparency obligations, with chatbots and deepfakes as concrete examples.PublishedNational guidanceDutch Data Protection Authority (AP), Algorithm Coordination Directorate20 Apr 2026
- Work agenda for coordinating AI and algorithm supervision 2026Names transparency and explainability as a priority in the year when transparency obligations actually come into force.PublishedNational guidanceDutch Data Protection Authority (AP), Algorithm Coordination Directorate26 Mar 2026
- Report AI & Algorithms Netherlands (RAN) - March 2026Substantiates with measurement data that transparency and explainability are falling short in Dutch practice, precisely the area for which the AP receives supervisory authority.PublishedNational guidanceDutch Data Protection Authority (AP)5 Mar 2026
- Response of the President of the Administrative Jurisdiction Division to the draft AI Regulation Implementation ActAdvises extending its own supervision to AI systems that the Division uses and that fall only under the transparency obligations of Chapter IV, without being high-risk.PublishedOpinionCouncil of State, Administrative Jurisdiction Division8 Jul 2026
What does this mean for you?
Provider+
Ensure your AI system clearly indicates that users are interacting with AI. For chatbots: display this at first contact. For generated content: provide metadata or labels indicating AI generation.
Deployer+
Inform your customers and employees when they interact with AI. This applies to chatbots, automated decision-making, and AI-generated content. Deepfakes must always be labelled.
SME / Startup+
Do you use an AI chatbot on your website? Ensure a clear notice: 'You are speaking with an AI assistant'. This is mandatory and easy to implement.
Public Sector+
Citizens have the right to know when they interact with AI. This applies to chatbots on government websites, automated decisions, and AI-assisted enforcement.
Compliance checklist
- AI interaction is clearly communicated
- AI-generated content is labelled
- Deepfakes are marked as artificial
- Emotion recognition is disclosed to affected persons
Want to save your progress? Create an account
Related recitals
Recital 132Primary
Certain AI systems intended to interact with natural persons or to generate content may pose specific risks of impersonation or deception irrespective of whether they qualify as high-risk or not. In c…
Recital 133Primary
A variety of AI systems can generate large quantities of synthetic content that becomes increasingly hard for humans to distinguish from human-generated and authentic content. The wide availability an…
Related tools
Related enforcement
Related blog posts
- →AI Act compliance: chatbots, voicebots & sentiment
- →AI content labeling rules: Article 50 AI Act, 2026
- →European Parliament votes on AI Act Omnibus: delays for high-risk AI and ban on nudifier apps
- →AI content transparency: EU code of practice
- →ChatGPT/Claude at work: EU AI Act obligations
- →Article 50 EU AI Act: provider vs deployer guide
Cross-references
Frequently asked questions
What transparency obligations apply to AI?+
Article 50 requires that users are informed when interacting with an AI system (such as chatbots), when content is AI-generated (deepfakes), and when emotion recognition or biometric categorisation is applied.
Do I need to disclose that my chatbot is AI?+
Yes, Article 50 requires that natural persons are informed they are interacting with an AI system, unless this is obvious from the circumstances.
What documentation does Article 50 of the AI Act require?+
Article 50 of the AI Act requires that relevant documentation is maintained as part of the compliance process. This may include technical documentation, instructions for use, logs or declarations of conformity, depending on the classification of the AI system.
Do I need to label deepfakes under the AI Act?+
Yes, Article 50(4) requires deployers publishing deepfakes (AI-generated or manipulated image, audio or video content) to clearly disclose that the content is artificially generated or manipulated. This applies to both visible labels and machine-readable marking.
When do the transparency obligations of Article 50 take effect?+
The transparency obligations of Article 50 take effect on 2 August 2026, together with most other provisions of the AI Act. The AI Office has published a Code of Practice to help organisations with implementation.
Do I need to tell users they are talking to a chatbot?+
Yes, Article 50(1) requires providers of AI systems intended for direct interaction with persons (such as chatbots) to ensure users are informed they are communicating with an AI system, unless this is obvious from the circumstances.
Does the labelling obligation also apply to AI-generated text on news websites?+
Yes, Article 50(4) states that AI-generated or manipulated text published to inform the public on matters of public interest must be labelled as artificially generated. This is specifically relevant for news websites and media organisations.