EU Proposes AI Act Delay to Boost Industrial Capacity
Brussels seeks to postpone high-risk AI requirements until 2027 while launching infrastructure initiatives to counter US and Chinese dominance.
The European Union has proposed a significant delay in the implementation of its landmark AI Act, signaling a strategic pivot from pure regulation toward industrial development. The move aims to reduce regulatory friction as the bloc struggles to maintain competitiveness against global tech leaders.
Under a "Digital Omnibus" package introduced in November 2025, the European Commission seeks to postpone the implementation of requirements for high-risk AI systems until December 2027. This adjustment follows the AI Act's entry into force on August 1, 2024, which occurred after the Council's formal adoption on May 21, 2024. The proposal reflects an effort to simplify the EU's broader digital framework to better support innovation.
The Regulatory Evolution
Europe's approach to artificial intelligence has evolved through several stages. The journey began in 2017 with the establishment of ethical principles, followed by a 2020 White Paper and a formal proposal for the AI Act in 2021. While the original framework focused on banning dangerous AI and strictly regulating high-risk systems, the rise of generative AI forced policymakers to redesign requirements to account for general-purpose AI models.
A Shift Toward Sovereignty
This regulatory pause coincides with a new push to build domestic technical capabilities. On June 3, 2026, the EU proposed the Cloud and AI Development Act (CADA). This initiative is designed to strengthen European cloud and AI infrastructure, coordinating investment to reduce the bloc's heavy dependency on foreign providers, primarily those based in the U.S. and China.
Analysis suggests that regulating AI in isolation is insufficient for global competition. The shift toward CADA indicates a realization that the EU must aggressively build its own computing infrastructure and talent pool to avoid falling behind in the global race.
Market Implications and Uncertainty
Despite the proposed delays, the transition period remains fraught with ambiguity. Regulatory uncertainty persists, and there is a possibility that the effective scope of the AI Act could ultimately prove far broader than many industry stakeholders initially anticipated.
For the industry, this means a temporary reprieve on high-risk compliance but a long-term requirement to align with a framework that is still being refined. The EU's strategy now rests on a delicate balance: maintaining its commitment to "trustworthy AI" while ensuring that its rules do not stifle the very innovation it hopes to foster through CADA.