AI Act implementation timeline
The EU’s Artificial Intelligence Act, published in July 2024, takes a risk- and capability-based approach to regulating AI systems and GPAI models like large language models. While many have seen early drafts and expert proposals (including ones I contributed to), we now have the official implementation guideline from the EU.
Just a reminder: The Act is rolled out over several years, most obligations apply from August 2026, with full effect in 2027.
The guidance includes:
🧭 Implementation timeline
⚖️ Rules on risk-based system classification
🧩 Provisions for GPAI and innovation support
📏 The role of harmonised standards (Article 40)
🧠 Codes of conduct and practices (Articles 56, 95, 96)
If you’re working on AI compliance, trust, or governance, this document is worth your time.