🇪🇺 EU AI Act Implementation Timeline

AI Act Enters into Force

The EU Artificial Intelligence Act was officially published in the Official Journal and entered into force, marking the beginning of the world's first comprehensive AI regulation.

Prohibited AI Systems Ban

The first substantive obligations came into effect, banning AI systems with unacceptable risk levels.

• 🚫 Social scoring systems by public authorities
• 🚫 Real-time biometric identification in public spaces
• 🚫 AI systems exploiting vulnerabilities
• 🚫 Subliminal techniques causing harm

GPAI Obligations Launch

General Purpose AI model obligations become legally binding, with the GPAI Code of Practice providing voluntary compliance guidance.

• 📋 Technical documentation requirements
• 📊 Transparency reports on capabilities & risks
• 📚 Training data summaries (including copyright)
• 🔒 Systemic risk models: additional safety measures
• 🏛️ EU AI Office becomes operational

High-Risk AI Systems & Full Enforcement

The AI Act becomes generally applicable with comprehensive obligations for high-risk AI systems and full enforcement powers.

• ⚠️ High-risk AI systems obligations (Annex III)
• 🔍 Member state enforcement authorities operational
• 🏖️ AI regulatory sandboxes established
• 💰 Full penalty regime in effect
• 🏷️ AI-generated content labeling requirements

Full Compliance Requirement

All AI systems, including legacy GPAI models and high-risk AI systems, must achieve full compliance.

• 🕐 Legacy GPAI models (pre-Aug 2025) must comply
• 🔧 High-risk AI systems with 3rd-party assessments
• 🗂️ Large-scale IT systems integration