The AI Regulation (EU AI Act, EU 2024/1689) is the worldβs first comprehensive regulatory framework for artificial intelligence. Published in July 2024, it takes a risk-based approach with proportional obligations.
Risk classification
- Unacceptable risk: prohibited systems (mass biometric recognition, social scoringβ¦)
- High risk: systems in critical sectors (healthcare, education, employment, infrastructureβ¦) β strict obligations
- Limited risk: chatbots, deepfakes β transparency obligations
- Minimal risk: no additional requirements
Obligations for high-risk systems
Risk management, data governance, technical documentation, transparency, human oversight, robustness and cybersecurity, and registration in the EU database.
ermine and EU AI Act
ermine manages the AI systems inventory, classifies their risk level, documents conformity requirements and keeps the technical file up to date for regulatory audits.