EU AI Act

The EU AI Act is the first comprehensive AI regulation, establishing risk-based requirements for AI systems sold or used in the European Union.

The EU AI Act creates a regulatory framework for artificial intelligence, classifying AI systems by risk level with corresponding requirements.

Risk categories: - Unacceptable Risk (Banned): Social scoring, real-time biometric ID in public - High Risk: Critical infrastructure AI, employment, education, law enforcement - Limited Risk: Chatbots, emotion recognition (transparency required) - Minimal Risk: Most AI applications (no restrictions)

High-risk AI requirements: - Risk management system - Data governance and quality - Technical documentation - Record-keeping and logging - Transparency to users - Human oversight - Accuracy, robustness, cybersecurity

Timeline: - Prohibited AI: 6 months after entry into force - High-risk: 24-36 months - Full application: By 2027

Why It Matters

The EU AI Act is the world's first comprehensive AI regulation, with fines up to €35 million or 7% of global revenue. It applies to any company whose AI systems are used in or affect people in the EU—regardless of where the company is headquartered. Organizations deploying high-risk AI must implement risk management systems, technical documentation, and human oversight requirements that are already being enforced for prohibited AI categories.

Key Points

Risk-based classification approach

High-risk AI has extensive requirements

Applies to AI used or sold in EU

Fines up to €35M or 7% global revenue

Phased implementation through 2027

Applicable Compliance Frameworks

GDPR ISO 27001

Related Terms

AI Governance

AI governance is the framework of policies, processes, and controls that ensure AI systems are developed and used responsibly, ethically, and in compliance with regulations.

ISO 42001

ISO/IEC 42001 is the international standard for AI management systems, providing a framework for organizations to responsibly develop and deploy artificial intelligence.

Algorithmic Accountability

Algorithmic accountability ensures that organizations can explain, justify, and take responsibility for the outcomes of automated decision-making systems.