EU fails to reach deal on weakened AI Act after 12-hour negotiation
EU countries and European Parliament lawmakers broke off talks Tuesday without agreement on changes to the bloc's artificial intelligence rules after 12 hours of negotiations. They plan to resume discussions next month.
The proposed modifications would loosen requirements in the AI Act, which took effect in August 2024. The European Commission included the changes in its Digital Omnibus package, designed to streamline digital regulations so European companies can compete with U.S. and Asian competitors.
A Cypriot official, speaking for the EU Council presidency, said reaching agreement with Parliament proved impossible. Dutch lawmaker Kim van Sparrentak criticized the stalled talks, saying: "Big Tech is probably popping champagne. While European companies that care about safety and did their homework now face regulatory chaos."
Where negotiations broke down
Talks stalled over exemptions for industries already subject to separate regulations. Some countries and lawmakers pushed to exclude sectors with existing product safety rules from AI legislation requirements.
The current AI Act imposes strict requirements for "high-risk" applications including biometric identification, utility supply, health assessments, creditworthiness decisions, and law enforcement.
The Digital Omnibus package also covers the General Data Protection Regulation, the e-Privacy Directive, and the Data Act. Privacy advocates and civil rights groups have criticized proposed changes across these regulations as concessions to large technology companies.
Negotiators are expected to reconvene in roughly two weeks.
Learn more: Explore AI for Government or the AI Learning Path for Policy Makers for deeper context on regulatory frameworks and policy development.
Your membership also unlocks:
