European Union · Mandatory regulation
EU AI Act, compliant
The world's first comprehensive AI law. Mandatory for any AI that affects EU citizens — with fines up to €35M or 7% of global revenue.
- Status
- In force — obligations phasing in
- High-risk obligations apply
- August 2026
- Maximum fines
- €35M / 7% of revenue
- Controls in Unorma
- 47
What it is
The European Union Artificial Intelligence Act, in plain language.
The EU AI Act takes a risk-based approach: AI systems are classified as prohibited, high-risk, limited-risk or minimal-risk, and obligations scale with the classification. High-risk systems — from CV screening to medical devices — face the full weight of the law: risk management, data governance, technical documentation, human oversight, and post-market monitoring.
The Act applies to providers who build AI and deployers who use it, wherever they are based, as long as the AI affects people in the EU. Prohibitions and AI literacy duties already apply; the high-risk regime lands in August 2026. Waiting is not a strategy — the documentation the Act demands takes months to build.
Who it applies to
Providers
You build or substantially modify AI systems placed on the EU market. You carry the heaviest obligations: conformity assessment, technical documentation, CE marking and registration.
Deployers
You use AI systems under your authority in the EU. You owe human oversight, input data quality, monitoring, and — for some systems — fundamental rights impact assessments.
Importers & distributors
You bring third-party AI to the EU market. You must verify the provider's compliance before the system reaches users.
Requirements
What EU AI Act actually asks of you.
47 controls in Unorma — each with plain-language guidance, action steps and the evidence you need.
Risk management — Art. 9
A continuous, documented risk management system across the AI lifecycle.
Data governance — Art. 10
Training, validation and test data must be relevant, representative and managed for bias.
Technical documentation — Art. 11
Comprehensive documentation proving compliance, kept current and audit-ready.
Transparency — Art. 13
Instructions for use and disclosures so deployers and users understand the system.
Human oversight — Art. 14
Systems designed so humans can effectively supervise, intervene and override.
Incident reporting — Art. 73
Serious incidents reported to market surveillance authorities on strict timelines.
EU AI Act in Unorma
From regulation text to audit-ready, in one platform.
Automatic risk classification
The AI Inventory wizard classifies each system as high, limited or minimal risk and determines whether you're a provider or deployer — then generates exactly the obligations that apply.
Six EU AI Act document templates
Model Cards, Risk Management Records, Human Oversight Procedures, Transparency Notices, Data Governance Records and Instructions for Use — AI-generated, pre-filled with your system data.
Art. 73 authority reports
When an incident happens, Unorma pre-fills the regulatory notification so you meet the deadline without panic.
Art. 4 AI literacy, proven
Eight built-in EU AI Act training modules with quizzes and certificates that auto-save to the Evidence Vault.
Track EU AI Act readiness from day one.
14-day free trial · No credit card · All 47 controls included
Cross-framework intelligence
Evidence for EU AI Act counts elsewhere too.
~100 cross-framework mappings mean work done once counts toward multiple frameworks automatically.