What the EU AI Act means for you, in plain English

The EU Artificial Intelligence Act is the first comprehensive, horizontal law for AI. If you live in the European Union, interact with an EU business, or use products made by EU companies, parts of it are likely to touch you.

The short version The AI Act groups AI by risk. A handful of uses are banned. A bigger group is 'high-risk' and has to meet strict rules. Some uses just have to be transparent. And there are new obligations for the big general-purpose models that power many of today's AI products.

Banned uses The Act bans a short list of especially harmful AI practices, including social scoring by governments, manipulative techniques that exploit people's vulnerabilities, real-time biometric identification in public spaces (with narrow exceptions), and untargeted scraping of face images.

High-risk AI This is the heart of the Act. AI used in credit, hiring, education, law enforcement, migration, critical infrastructure, and several other sensitive areas has to be risk-managed, tested, documented, monitored, and overseen by a human.

What changes for you You should see clearer disclosures when interacting with AI, labels on AI-generated content, the ability to ask for human review of important AI decisions, and a path to file a complaint with a market surveillance authority in your country.

What to watch The Act phases in between 2024 and 2027. Bans started first, general-purpose AI rules follow, and the bulk of the high-risk rules apply later. If a business you use operates in the EU, expect to see new AI notices and choices rolling out over this window.