EU Nations Approve Deal to Ease AI Act Restrictions
BRUSSELS — European Union member states provisionally approved a deal to ease artificial intelligence restrictions under the bloc’s landmark AI Act, according to AI Business and EU Today. The agreement still awaits formal European Parliament endorsement.
The agreement, described as an AI Act simplification package, was reached by EU negotiators.
The deal would modify the EU’s original approach to AI regulation. The AI Act, finalized in 2024 as the world’s first comprehensive AI law, established a risk-based framework imposing strict requirements on high-risk AI systems, including transparency obligations, conformity assessments and prohibitions on certain uses of the technology.
The simplification package comes as the EU faces an August 2026 deadline for full enforcement of the AI Act’s provisions. The package follows industry concerns that overly prescriptive rules could hamper innovation and put EU-based companies at a competitive disadvantage.
Impact on US Companies
The regulatory changes carry direct implications for major American AI companies. OpenAI, Google, Meta, Microsoft and Anthropic all maintain significant operations in the European market and face compliance obligations under the AI Act.
A more permissive regulatory environment in Europe could reduce compliance costs for US firms expanding into or operating within EU markets. The changes may also influence ongoing debates in Washington over federal AI legislation, as US lawmakers have closely monitored the EU’s regulatory approach as a potential model for domestic policy.
US technology companies have lobbied extensively against what they characterized as burdensome provisions in the original AI Act, arguing that rigid compliance requirements could slow the deployment of beneficial AI applications and create barriers to market entry.
What Comes Next
The provisional agreement must still clear the European Parliament before taking effect. Deals securing member state backing typically proceed to Parliament approval, though the legislative body could seek modifications or delay the timeline.
The simplification effort reflects ongoing policy debate within the EU between its goal of setting global standards for responsible AI development and pressure from industry leaders and some member states to avoid regulation that could affect the bloc’s competitiveness.
Details of which specific provisions would be eased under the simplification package were not immediately available from the published reports.
