EU AI Act
The EU Artificial Intelligence Act (2024/1689), a risk-based regulation of AI systems, with obligations phasing in from 2025.
Overview
The AI Act prohibits certain AI practices, sets requirements for high-risk AI systems including risk management, data governance, human oversight and documentation, and imposes duties on providers of general-purpose AI models. It is overseen by the EU AI Office and national authorities.
As regulation it is binding and specific, and it is the layer most directly mapped to controls. The authorities most closely connected to it are European AI Office.
Check the official sources below for legal scope, applicable dates and exceptions. This entry is an explanatory reference, not legal advice.
Why it matters
It is the first comprehensive AI law, and it reaches AI used in regulated decisions.
Who it affects
Providers and deployers of AI systems in the EU.
Where it sits
Ten linked steps connect where you operate to the proof that you complied.
- 01Jurisdiction
- 02Industry
- 03Activity
- 04Authority
- 05Regulation
- 06Requirement
- 07Guidance
- 08Control
- 09Evidence
- 10Enforcement
Connections
The terms most closely tied to EU AI Act. Select one to open it.
What to do about it
- 1Confirm whether you are in scope: who it covers, which activities and where.
- 2Break it into testable requirements.
- 3Map each requirement to a control and an owner.
- 4Track amendments, implementing rules and guidance.
- 5Keep sealed evidence for every control run.
From requirement to evidence.
AI Act obligations are mapped to human oversight, documentation and monitoring controls.
Request early access →