Glossary
EU AI Act
The EU AI Act is the European Union's law regulating artificial intelligence. It sorts AI uses by risk: banning some, putting strict duties on high-risk systems, and requiring transparency for chatbots and AI-generated content. Its main rules phase in from 2025 to 2028.
The AI Act is the EU’s attempt to do for artificial intelligence what the GDPR did for personal data: one rulebook for the whole Union. It entered into force in August 2024, but it applies in stages. The first bans applied from February 2025, the duties for makers of general-purpose models from August 2025, and the transparency rules from August 2026. In July 2026 an amending law, Regulation (EU) 2026/1744, pushed the high-risk rules back: to December 2027 for uses like hiring and credit, and to August 2028 for AI built into products such as medical devices.
The core idea is a risk ladder. A few practices are banned outright (think social scoring). High-risk systems, like AI used in hiring, credit or medical devices, face strict requirements. And for everyday tools the keyword is transparency: you must be told when you’re talking to an AI, and AI-generated or manipulated content like deepfakes must be disclosed. AI tools that generate content and were already on the market before August 2026 have until 2 December 2026 to mark what they make as AI-generated. From that same date, AI tools that make sexual deepfakes of real people without their consent are banned, whether they were built for it or simply lack safeguards against it. Makers of general-purpose models such as the ones behind ChatGPT, Claude and Gemini have their own documentation and transparency duties, and the European Commission’s AI Office supervises them directly.
For most users and small businesses, the practical effect is reassuring rather than burdensome: clearer labels, more honest disclosure, more accountable vendors. If you’re choosing which assistant to rely on, our AI chooser compares the major options.
This page explains the law’s shape, not your obligations. For those, the European Commission’s AI Act pages and national authorities are the source.
Where you’ll meet this
- “AI-generated” labels appearing on images, video and audio across platforms
- Vendor compliance and transparency pages referencing the AI Act
- The European Commission’s AI Act portal and national implementation pages