Skip to main content
European UnionIN FORCE

EU AI Act: Regulation (EU) 2024/1689

Issuer: European Parliament & Council

Date: 1 AUG 2024

Status: IN FORCE in European Union, as of 29 JUL 2026

World's first horizontal AI law; four-tier risk pyramid; extraterritorial (Art. 2).

One single rulebook for all 27 member states. It regulates what AI is used for, not the technology itself: prohibited practices are banned outright, high-risk uses carry the full conformity stack, limited-risk systems owe transparency, and ~85% of AI (minimal risk) faces no new duties. Applies to any provider or deployer whose AI output is used in the EU.

What it actually requires (5 provisions)

  • Art. 5 prohibitions: social scoring, manipulation, workplace/school emotion recognition, most police real-time facial recognition (in force 2 Feb 2025)
  • Annex III high-risk: recruitment, credit scoring, education, essential services, law enforcement, biometrics
  • Art. 50 transparency: chatbot disclosure, machine-readable synthetic-content marking, deepfake labels
  • Ch. V GPAI: documentation, copyright policy, training-data summary, systemic-risk tier >10²⁵ FLOPs
  • Art. 99 to 101 penalties: up to €35M / 7% of worldwide turnover for prohibited practices

How its status moved

  1. PUBLISHEDOJ L, 12 Jul 2024
  2. IN FORCE1 Aug 2024
  3. CURRENTPhased application: prohibitions live, high-risk duties 2 Dec 2027 / 2 Aug 2028

Sources (2)

The rest of the European Union stack

5 more instruments in this jurisdiction, each with its own status, provisions and sources.

Where this sits in the wider picture

Knowing the instrument is step one. Complying with it is the job.

The programs teach the work that follows a rule like this one: classification calls, conformity assessment, filings, documentation, and the judgment to defend every decision.

VERIFIED 29 JUL 2026. Every fact on this page is drawn from the sources listed above and dated to the day it was checked. Study aid, not legal advice: the official texts are always authoritative.