European Commission Rejects Tech Firms' Call to Delay AI Act

The European Commission has firmly rejected calls from major tech firms to delay the implementation of its AI Act, reaffirming its commitment to the legislation's timeline. This decision comes despite an open letter from over 100 CEOs, including those from Alphabet, Meta, Mistral AI, and ASML, urging a two-year pause. These companies argued that the new rules could hinder Europe’s competitiveness in the rapidly evolving global AI sector.

The AI Act, which introduces a risk-based framework to regulate AI systems, classifies applications under various levels of risk. Certain use cases deemed to pose "unacceptable risk," such as cognitive behavioral manipulation or social scoring, are outright banned. High-risk applications, including those involving biometric surveillance, education, and employment, are subject to strict requirements, including registration and compliance with safety, transparency, and quality standards. AI tools deemed "limited risk," such as chatbots, will face lighter regulatory burdens, primarily focused on transparency obligations.

The European Commission's spokesperson, Thomas Regnier, emphasized the EU's unwavering stance on the AI Act, stating, "There is no stop the clock. There is no grace period. There is no pause." This firm response underscores the EU's determination to proceed with the implementation of the AI Act as planned, despite the industry backlash.

The EU began implementing the AI Act in stages last year, with the full set of rules expected to take effect by mid-2026. The Commission dismissed the notion that a delay would benefit the European AI ecosystem, suggesting instead that clarity and regulatory certainty would foster innovation and growth. This decision is likely to have significant implications for tech companies operating in the EU, as they will need to comply with the new regulations or risk facing penalties.

Provisions of the law began taking effect in February of 2025. Obligations for general-purpose AI models will begin in August 2025, while the strictest rules that target high-risk AI systems will apply starting in August 2026. The Commission acknowledged the challenges raised by stakeholders and is pursuing a parallel effort to streamline broader digital regulations in the EU. According to Regnier, a proposal to simplify and reduce administrative reporting obligations is expected by the end of 2025. However, this initiative is separate from the AI Act and is not intended to delay or weaken its core provisions.

“We understand that small companies have different capacities, and we want to ensure rules are proportionate,” Regnier said. “But this does not mean pausing the AI Act or rewriting the legal timelines already in place.” The bloc sees its legal frameworks as a template that could influence global standards, much as its General Data Protection Regulation (GDPR) has done in the privacy and data protection industry. The AI Act also introduces fines reaching up to €35M or 7% of global turnover, depending on the severity of non-compliance or violation.