The European Union’s ambitious AI Act has achieved a significant milestone as Member State representatives voted to confirm the final text of the draft law.
The AI Act, designed as a risk-based plan for regulating various AI applications, underwent rigorous negotiations and faced opposition from some Member States, with France expressing concerns about potential legal limits on the growth of domestic AI startups.
The political agreement reached in December set the stage for turning agreed-upon positions into a final compromise text. Today’s Coreper vote, where all 27 ambassadors of EU Member States unanimously backed the final text, signifies a crucial step toward the adoption of the regulation. The AI Act outlines prohibited uses of AI, establishes governance rules for high-risk applications, and imposes transparency requirements on specific AI applications, such as chatbots.
📝 Signed!
— Belgian Presidency of the Council of the EU 2024 (@EU2024BE) February 2, 2024
Coreper I Ambassadors confirmed the final compromise text found on the proposal on harmonised rules on artificial intelligence (#AIAct).
The AI Act is a milestone, marking the 1st rules for AI in the 🌍, aiming to make it safe & in respect of 🇪🇺 fundamental rights. pic.twitter.com/QUe2Sr89A5
The successful vote alleviates concerns about potential derailment of the regulation, especially considering opposition led by France and lingering doubts from Germany and Italy.
The European Parliament will now take the baton, where lawmakers are expected to provide a final vote on the compromise text. Despite potential opposition, the widespread support from Member States suggests that the EU’s flagship AI Act is likely to be adopted in the coming months.
Once adopted, the AI Act will enter into force 20 days after publication in the EU’s Official Journal. A tiered implementation period will follow, with a grace period of six months before a list of prohibited uses begins to apply, likely around the fall. Rules on foundational models, considered general-purpose AIs, will not apply until 2025, allowing for a phased and careful rollout of the comprehensive AI regulations.
The European Commission has taken proactive steps by establishing an AI Office to oversee compliance, particularly for powerful foundational models posing systemic risks.
Additionally, recent measures have been announced to support the growth of homegrown AI developers, including the adaptation of the EU’s supercomputing network to facilitate generative AI model training. The AI Act is a crucial component of the EU’s strategy to navigate the challenges and opportunities presented by artificial intelligence.