With less than a month before key provisions of the European Union’s AI Act take effect, major tech companies and some political leaders are calling for a delay. Industry groups representing U.S. giants such as Alphabet (GOOGL.O) and Meta (META.O), as well as European players including Mistral and ASML (ASML.AS), have urged the European Commission to postpone the implementation by several years, citing a lack of clarity and readiness.
READ ALSO: European Foreign Ministers Set To Toughen Action Against Russia
The EU’s landmark AI Act, passed last year after prolonged negotiations among member states, is being rolled out in stages. However, some of the most significant rules—those targeting general purpose AI (GPAI) models—are set to come into force on August 2. These rules apply to foundational models developed by companies like Google, Mistral, and OpenAI.
Under the new requirements, companies must comply with transparency obligations such as preparing technical documentation, respecting EU copyright laws, and providing detailed summaries of the content used to train their AI systems. In addition, firms will need to test their models for bias, toxicity, and robustness before release. Models classified as high-impact or posing systemic risk face even tougher mandates, including adversarial testing, risk mitigation, incident reporting to the European Commission, and disclosures on energy consumption.
AI developers argue that compliance with the AI Act will bring significant costs and operational burdens especially for those building the models themselves. Beyond financial concerns, many companies say they are in the dark about how to meet the Act’s requirements due to the absence of formal guidance. The AI Code of Practice, meant to help firms interpret the rules, missed its expected release date of May 2.
“To address the uncertainty this situation is creating, we urge the Commission to propose a two-year ‘clock-stop’ on the AI Act before key obligations enter into force,” said an open letter signed by 45 European companies. The letter also called for a simplification of the rules to avoid stifling innovation, particularly in Europe, where firms often lack the legal resources available to larger U.S. competitors.
So far, the European Commission has not confirmed whether it will delay enforcement. However, EU tech chief Henna Virkkunen has pledged to release the long-awaited AI Code of Practice before August. Some political figures, including Swedish Prime Minister Ulf Kristersson, have labeled the current rules “confusing” and echoed calls for a pause.
“A bold ‘stop-the-clock’ intervention is urgently needed to give AI developers and deployers legal certainty, as long as necessary standards remain unavailable or delayed,” said CCIA Europe, a tech lobbying group.
The European Commission has not yet responded to requests for comment yet.