The European Commission is sticking to its planned timetable for the AI Act, even as tech startups and major companies ask for a multi‐year delay. Speaking from Brussels on Friday, spokesperson Thomas Regnier made it clear: “There is no grace period. There is no pause.”
The primary regulations kicked in back in February 2025. Additional rules for large AI models are set to roll out this August, with even stricter requirements for high‑risk systems coming in August 2026. These staged deployments aim to balance innovation with necessary oversight.
To ease the compliance burden on smaller enterprises, the EU is relaxing certain obligations, such as reporting requirements. While the Act has drawn worldwide attention for its ambitious reach, critics warn that its complex framework could result in significant costs and a patchwork of rules that are difficult to navigate.
Part of the strategy is to bolster the EU’s standing in a competitive AI landscape dominated by US companies. Despite OpenAI laying out an extensive list of demands in its so‑called “Economic Blueprint” for Europe, the organisation has offered few firm commitments, raising concerns that the region could end up serving mainly as a market and infrastructure provider for US-led AI innovations.
A coalition of over 45 companies—including ASML, Airbus, and Mistral AI—has called for a two‑year delay in implementing the Act. In a letter to Commission President Ursula von der Leyen, these businesses argued that the proposed rules for powerful AI models could stifle innovation. Names like Mercedes‑Benz, Lufthansa, BNP Paribas, Siemens Energy, and Black Forest Labs joined the criticism, with Meta and Alphabet already dismissing the delayed code of conduct as impracticable.
Despite these appeals, the new rules are set to take effect this August. Advocacy groups such as General Catalyst are closely monitoring the situation, with industry players like SAP and Spotify lending their support to the broader discussion, even if they didn’t sign the letter themselves.
