New Milestone Achieved in AI Governance

EU member states have unanimously voted for the Artificial Intelligence Act forward, setting the stage for changes in global AI regulation.

This decision, endorsed by all EU member states in COREPER, marks a crucial step forward despite recent concerns voiced by key stakeholders. European Commissioner Thierry Breton has commended the agreement for its balanced approach, fostering trust and innovation and ensuring the protection of citizens’ rights.

Now, all eyes turn to the European Parliament for the next critical steps, with key votes anticipated in February and April. The path ahead involves a phased implementation approach, with carefully mapped out key dates for various provisions.

Following extensive negotiations, a political consensus was reached on the EU Artificial Intelligence Act on December 8, 2023, likely establishing it as the world’s most comprehensive AI regulation upon implementation.

🗓️ Key Dates:

• Entry into Force: 20 days after publication in the official journal of the EU.
• Entry into Application: 24 months after entry into force, with exceptions for specific provisions.
• 6 Months After: Prohibitions on unacceptable risk AI.
• 12 Months After: Obligations on providers of general-purpose AI models and appointment of member state competent authorities.
• 18 Months After: Commission implementing act on post-market monitoring.
• 24 Months After: Obligations on high-risk AI systems listed in Annex III, penalties enforcement by member states, and establishment of AI regulatory sandboxes.
• 36 Months After: Obligations for high-risk AI systems not prescribed in Annex III.
• By The End of 2030: Obligations for certain AI systems as components of large-scale IT systems.

The European Commission’s role includes leading on delegated acts and establishing an AI Office to oversee compliance, demonstrating a commitment to effective AI governance.