The AI Act: Striking a Deal & Beyond
summary
European lawmakers claim the EU AI Act targets only concrete high-risk uses — those risking health, safety or fundamental rights — not whole sectors, and it includes concrete criteria for legal certainty. Negotiations took more than two years in Parliament with a final 36 hours of negotiations to seal the deal. The text adds a research and open-source exemption (applies until a product is placed on the market except for some powerful systems) and a two-tier approach for generative AI; phased enforcement runs from 6 months to 3 years.
AI-written summary · about
chapters (10)
- 0:00Introduction
- 1:55AI Act negotiations
- 3:43Safety versus innovation
- 5:30High-risk use cases
- 6:54After the AI Act deal
- 8:22Europe’s AI vision
- 10:10Investing in innovation
- 11:55Political leadership
- 13:43Global AI standards
- 14:50Conclusion
related talks
26mThe European AI Act: Challenges and Opportunities
28mAI Act - What is important to know from a developers perspective?
01h 16mTHE AI ACT: WHERE ARE WE, AND WHERE ARE WE GOING?
01h 08mTHE GOVERNANCE OF AI: CONVERGENCE OR DIVERGENCE?
01h 17mGoverning AI in Europe Advancing Fundamental Rights...
01h 19mPower, Convenience, and Prestige in the Governance and Legal Regulation of AI
01h 08mAI Act in Action Compliance, Oversight, and Enforcement
35mThe European AI Act: a data scientist’s perspective