EU AI Act — Article 50 (Transparency Obligations)
The transparency-obligations article of the EU’s AI Act, covering four situations where a person interacting with or exposed to an AI system’s output must be told so: chatbot disclosure, synthetic-content marking, deepfake/public-interest-text disclosure, and emotion-recognition/biometric-categorization notice.
License: Official EU legislative text, freely reusable per Commission Decision 2011/833/EU
Key points
- A provider of a chatbot or other AI system that interacts directly with a person must ensure that person is informed they’re talking to an AI, unless that’s already obvious to a reasonably informed person from context, disclosed no later than the first interaction. See AI Content Marking.
- A provider of a system generating synthetic audio, image, video, or text must mark that output in a machine-readable format detectable as AI-generated or manipulated, accounting for technical feasibility. See AI Content Marking.
- A deployer of a deepfake, or of AI-generated text published to inform the public on a matter of public interest, must disclose that artificial origin — except once the content has undergone genuine human editorial review with a named party taking responsibility for it, or for purely assistive edits that don’t substantially alter deployer-supplied input. See AI Content Marking.
- A deployer of an emotion-recognition or biometric-categorization system must inform the people exposed to it that the system is operating, subject to EU data protection law.