EU countries and European Parliament lawmakers failed to agree on revised artificial intelligence rules after 12 hours of negotiations on Tuesday, April 29, according to Reuters. Talks will resume next month.
The proposed changes to the AI Act, which entered into force in August 2024 with enforcement phased across 2025 and 2026, are part of the European Commission’s Digital Omnibus package. That package aims to simplify digital regulations so European businesses can compete with U.S. and Asian rivals, according to CNA.
The Sticking Point
Negotiations broke down over a specific demand: some countries and lawmakers insisted that industries already subject to sectoral regulations, such as product safety rules, should be exempt from the AI Act entirely, The Indian Express reported. The AI Act currently imposes stricter requirements on “high-risk” uses of AI in areas like biometric identification, utilities, health, creditworthiness, and law enforcement.
“It was not possible to reach an agreement with the European Parliament,” a Cypriot official told Reuters. Cyprus currently holds the rotating EU Council presidency.
Industry and Civil Society React
Dutch lawmaker Kim van Sparrentak criticized the outcome in a statement: “Big Tech is probably popping champagne. While European companies that care about safety and did their homework now face regulatory chaos,” according to CNA.
The Digital Omnibus package also encompasses the General Data Protection Regulation (GDPR), the e-Privacy Directive, and the Data Act. Privacy activists and civil rights groups have criticized the proposed changes as concessions to Big Tech, according to Reuters.
People with direct knowledge of the negotiations said the next round of discussions will likely take place in two weeks, according to CNA.
The Compliance Gap for Agent Builders
For companies deploying AI agents across EU member states, the failed negotiations create a specific problem: the current AI Act is the strictest AI regulation in the world, but the proposed revisions were designed to relax it. With no agreement on what gets relaxed and what stays, enterprises face the full weight of existing rules while also lacking clarity on what the rules will look like in six months.
Agent deployments in high-risk sectors (healthcare, financial services, law enforcement) remain subject to the Act’s most stringent requirements. Whether industries with existing regulatory frameworks get carved out of AI-specific obligations remains unresolved until the next round of talks.