EU AI Act

European Union AI

Status

In force since 1 Aug 2024, applying in stages. The AI Omnibus (in force 27 Jul 2026) moved stand-alone high-risk rules to 2 Dec 2027 and product-embedded high-risk rules to 2 Aug 2028.

Who it applies to

Providers placing AI systems on the EU market, deployers using AI in the EU, and providers or deployers outside the EU whose AI output is used in the EU.

Size thresholds

None for prohibitions and transparency duties. Obligations depend on the AI system's risk level and your role (provider, deployer, importer, distributor).

Regulator

European AI Office (general-purpose AI models and some AI in very large platforms) and national market surveillance authorities

Breach or incident reporting

Providers of high-risk AI systems must report serious incidents to market surveillance authorities (Art. 73).

Data protection officer

N/A (AI literacy measures required under Art. 4, as amended by the AI Omnibus)

Local representative

Being verified

Maximum penalty

Up to EUR 35m or 7% of worldwide turnover for prohibited practices; up to EUR 15m or 3% for other obligations, including high-risk and transparency duties.

Facts marked “Being verified” are still being confirmed and are not given as answers.

What you must do

• Stop any prohibited AI practice (banned since 2 Feb 2025).
• Tell people when they interact with an AI system; mark AI-generated content (Art. 50, from 2 Aug 2026).
• Classify each AI system you build or use by risk level and document the result.
• High-risk providers: risk management, data governance, logging, human oversight, conformity assessment before 2 Dec 2027.
• General-purpose AI model providers: technical documentation and a training-content summary (since 2 Aug 2025).

Recent changes

AI Omnibus (Reg. 2026/1744) in force 27 Jul 2026: high-risk deadlines deferred; transparency duties kept at 2 Aug 2026 with a marking grace period to 2 Dec 2026 for systems already on the market; two new prohibited practices added.

Sources

• Primary source: https://natlawreview.com/article/eu-digital-omnibus-ai-enters-force
• Second source: https://labs.cloudsecurityalliance.org/research/csa-research-note-eu-ai-act-article50-watermarking-deadline/
• Regulator: European AI Office (general-purpose AI models and some AI in very large platforms) and national market surveillance authorities

Information only, not legal advice. Laws change; confirm with the regulator or a qualified lawyer before relying on this page. Spotted an error? Email [email protected].