# What is the EU AI Act?

The EU AI Act is the European Union's law on artificial intelligence, in force since 1 August 2024. It sorts AI systems into four risk tiers, bans the top one, and asks more of a system the higher its tier.

## How it works

The Act is [Regulation (EU) 2024/1689](https://eur-lex.europa.eu/eli/reg/2024/1689/oj/eng). It sorts AI systems into four risk tiers: unacceptable (banned), high, limited (transparency duties) and minimal. General-purpose AI models have duties of their own.

Fines for banned practices reach EUR 35 million or 7% of worldwide turnover.

## Timeline

The Act has been in force since 1 August 2024 and applies in stages. Bans applied from February 2025, and duties for general-purpose AI models from August 2025.

Most high-risk duties were first set for August 2026. An amendment in force since July 2026, Regulation (EU) 2026/1744, moved them to December 2027, and to August 2028 for AI built into regulated products such as medical devices. Check the current text before you plan around a date.

## What it asks of a team

In general terms:

- **Know your risk tier.** Work out where each of your AI systems falls. What applies depends on it.
- **Document.** Record what the system does, how it was tested and what controls it has.
- **Monitor.** Watch the system in use and keep records. See [audit logs](/glossary/audit-logs) and [AI observability](/glossary/ai-observability).

This page is not legal advice. For the wider picture, see [AI compliance](/glossary/ai-compliance).

## With White Circle

[Analyze](/products/analyze) has audit logs for compliance and incident review, and policies can differ by the user's location or jurisdiction.

