Reference · Glossary

Article 50 & EU AI Act Glossary

Plain definitions of the terms used across Article 50 compliance, with the provision each comes from where there is one. Follow the links for the full explainer on each topic.

Scan for free

A–Z

AI Liability Directive: A proposed EU directive on civil liability for harm caused by AI. The Commission announced its withdrawal in February 2025 and formally withdrew it in October 2025; it never became law. Read more →
AI system: Software that infers from the input it receives how to generate outputs such as content, predictions, recommendations or decisions — for example a chatbot, an image generator or a voice cloning tool. Read more →
Article 50 guidelines: Guidelines the European Commission adopted in July 2026 on how Article 50 should be applied. They set out the Commission's interpretation but aren't binding law. Read more →
C2PA Content Credentials: An open standard for embedding provenance information — including that content was AI-generated — into image, video and audio files. One common way to implement machine-readable marking. Read more →
Deep fakeArt. 3(60): AI-generated or manipulated image, audio or video content that resembles existing persons, objects, places, entities or events and would falsely appear to a person to be authentic or truthful. Read more →
DeployerArt. 3(4): A person or organisation using an AI system under its own authority, except in the course of a personal non-professional activity. A business that embeds a vendor's chatbot on its site is a deployer. Read more →
Digital Omnibus on AIRegulation (EU) 2026/1744: Regulation (EU) 2026/1744, which amended the AI Act from 27 July 2026. Among other changes, it gave generative systems already on the market until December 2, 2026 to comply with Article 50(2), and extended the lower fine cap to small mid-caps. Read more →
Editorial review exceptionArt. 50(4): AI-generated public-interest text doesn't need a disclosure if it has undergone human review or editorial control and a person or organisation holds editorial responsibility for its publication. Both conditions must be met. Read more →
Fine tier for Article 50Art. 99(4): Breaches of any Article 50 transparency obligation can be fined up to €15 million or 3% of total worldwide annual turnover for the preceding financial year, whichever is higher. Read more →
Interaction disclosureArt. 50(1): The requirement that people are informed they are interacting with an AI system, unless that's obvious to a reasonably well-informed, observant and circumspect person. The duty is on the provider. Read more →
IPTC digital source type: A photo-metadata field recording how an image or video was made. The value trainedAlgorithmicMedia marks content created with generative AI; compositeWithTrainedAlgorithmicMedia marks real media edited with it. Read more →
Lower fine cap (SMEs and small mid-caps)Art. 99(6a) (small mid-caps): For SMEs, including startups, and for small mid-cap enterprises, an Article 50 fine is capped at the lower of the flat amount and the turnover percentage, not the higher. Read more →
Machine-readable markingArt. 50(2): A signal attached to AI-generated output that software can detect — embedded metadata, a watermark or similar — so the content can be identified as artificially generated or manipulated. Required of providers of generative AI systems. Read more →
Output used in the EUArt. 2(1)(c), Recital 22: The test that brings providers and deployers established outside the EU into scope when their AI system's output is used in the EU. The recitals describe it as output intended to be used in the EU. Read more →
Personal non-professional useArt. 2(10): The AI Act doesn't apply to the obligations of deployers who are natural persons using AI systems in the course of a purely personal, non-professional activity. Read more →
Product Liability Directive (revised)Directive (EU) 2024/2853: Directive (EU) 2024/2853, which treats software including AI systems as a product with strict liability for manufacturers. Member States must transpose it by December 9, 2026; it applies to products placed on the market after that date. Read more →
ProviderArt. 3(3): A person or organisation that develops an AI system, or has one developed, and places it on the market or puts it into service under its own name or trademark, whether for payment or free of charge. Read more →
Public-interest textArt. 50(4): AI-generated or manipulated text published with the purpose of informing the public on matters of public interest. Deployers must disclose it, subject to the editorial review exception. Read more →
Synthetic content: Audio, images, video or text generated or manipulated by an AI system rather than captured or written by a person. Article 50 requires it to be marked by the provider, and — for deep fakes and public-interest text — disclosed by the deployer. Read more →
Transition period for generative systemsArt. 111(4): Providers of generative AI systems already on the market before August 2, 2026 have until December 2, 2026 to comply with the Article 50(2) marking requirement. Read more →
Transparency obligationArt. 50: Under Article 50, a duty to make AI involvement visible to the people affected — telling them they're interacting with an AI, or that content is artificially generated or manipulated. Separate from the Act's high-risk system rules. Read more →

Check your site automatically

Article50.io is an automated scanner for the Article 50 transparency rules of the EU AI Act. It scans a website for potential Article 50 transparency gaps and generates a report of what the scan found and what it could not check, with remediation guidance, implementation instructions and disclosure wording you can paste in.

The free scan shows your single most severe finding in about 30 seconds — no signup, public pages only.

Scan for free

A–Z

Check your site automatically

Related Article 50 resources