EU AI Act
Regulation on Artificial Intelligence
The world’s first comprehensive AI law. Establishes risk-based classification for AI systems with strict prohibitions, conformity assessments for high-risk AI, transparency obligations, and GPAI model requirements. Penalties up to EUR 35M or 7% of global turnover.
113
Articles
13
Annexes
4
Risk Tiers
€35M
Max Fine
Implementation Timeline
1 August 2024
AI Act enters into force
2 February 2025
Prohibited AI practices ban takes effect
2 August 2025
GPAI model provider obligations apply
27 July 2026
Digital Omnibus on AI (Regulation (EU) 2026/1744) enters into force, postponing high-risk deadlines
2 August 2026
General date of application (e.g. Article 50 transparency, penalties)
2 December 2026
New prohibitions on AI-generated intimate imagery and CSAM apply; AI-content marking (Art. 50(2)) for systems already on the market
2 December 2027
High-risk AI system requirements apply (Annex III use cases)
2 August 2028
High-risk requirements for AI in regulated products (Annex I)
Who's In Scope
AI Providers
- Developers placing AI systems on the EU market
- GPAI model providers (general-purpose AI)
- Open-source AI developers (limited exemptions)
AI Deployers
- Organisations using AI systems in the EU
- Public bodies deploying AI in decision-making
- Private sector operators of high-risk AI
High-Risk Domains
- Biometric identification and categorisation
- Critical infrastructure management
- Credit scoring and employment decisions
- Education and vocational training
- Law enforcement and migration management
- Administration of justice and democratic processes
Key Requirements
Prohibited AI Practices
Bans social scoring, real-time remote biometric identification in public spaces (with limited exceptions), manipulation techniques exploiting vulnerabilities, and emotion recognition in workplace and education settings.
High-Risk AI Obligations
Risk management system, data governance, technical documentation, record-keeping, transparency to users, human oversight, and requirements for accuracy, robustness, and cybersecurity.
Transparency Obligations
All AI systems interacting with persons must disclose they are AI. Deepfakes and AI-generated content must be labelled as such.
GPAI Model Requirements
Technical documentation, copyright policy compliance, compute reporting for systemic risk models, adversarial testing, and serious incident reporting.
Fundamental Rights Impact Assessment
Deployers of high-risk AI in public sector and certain private contexts must assess impact on fundamental rights before deployment.
Post-Market Monitoring & Incidents
Providers must establish a post-market monitoring system and report serious incidents to market surveillance authorities.
Penalties & Enforcement
Prohibited AI violations
EUR 35M or 7% global turnover
Whichever is higher. Applies to banned practices under Art. 5.
Other non-compliance
EUR 15M or 3% global turnover
Applies to high-risk obligations, GPAI requirements, and most other provisions.
False information to authorities
EUR 7.5M or 1% global turnover
Supplying incorrect, incomplete, or misleading information to authorities.
How Orizon Can Help
Compliance Management
We provide structured preparation and compliance support for EU AI Act.
Official Sources
- EUR-Lex — Regulation (EU) 2024/1689https://eur-lex.europa.eu/eli/reg/2024/1689/oj
- Regulation (EU) 2026/1744 (Digital Omnibus on AI)https://eur-lex.europa.eu/eli/reg/2026/1744/oj
- AI Act Explorerhttps://artificialintelligenceact.eu/
- European Commission — AI Acthttps://digital-strategy.ec.europa.eu/en/policies/regulatory-framework-ai
Data Verification
Last verified:
Assess Your Readiness
Schedule a consultation with our regulatory experts to assess your EU AI Act compliance posture and build a practical roadmap.