The European Commission is considering pushing back parts of the EU’s Artificial Intelligence Act after sustained pressure from businesses and the Trump administration, according to media reports, Azernews reports, citing Tribune.
The Commission confirmed that ‘a reflection’ was ‘still ongoing’ on possible delays, following a Financial Times report that said Brussels was assessing changes intended to ease requirements for major tech companies.
The act, the world’s first broad legal framework for governing AI, took effect in August 2024, though many of its rules have not yet kicked in.
Most requirements for companies building high-risk AI systems that ‘pose serious risks to health, safety or fundamental rights’ are scheduled to apply starting in August 2026, or even a year later.
The Financial Times reported that the Commission is weighing a one-year ‘grace period’ for companies violating the strictest AI rules.
Under an internal document cited by the Financial Times, generative AI providers with products already on the market before the rules take effect could receive a one-year pause ‘to provide sufficient time . to adapt their practices within a reasonable time without disrupting the market.’
The report said the Commission is also weighing postponing fines for breaches of the new AI transparency rules until August 2027 to ‘provide sufficient time for adaptation of providers and deployers of AI systems.’
The MLex news organisation reported that Brussels is likewise exploring more flexible, less prescriptive monitoring requirements for developers of high-risk AI systems.
The proposals may still change before their expected Nov. 19 release, after which EU states and the European parliament must sign off, the Guardian reported.
Brussels has faced pressure from the Trump administration to soften tech rules, with the US president warning of tariffs on countries whose regulations ‘harm or discriminate against American technology.’
Meta, the parent company of US social media firms like Facebook and Instagram, has already declined to sign the commission’s code of practice for general-purpose AI models. ‘Europe is heading down the wrong path on AI,’ wrote Joel Kaplan, arguing the code creates ‘legal uncertainties’ and goes ‘far beyond the scope of the AI Act.’
Dozens of European firms have also called for a two-year halt to the act, arguing it is needed for ‘reasonable implementation’ and ‘further simplification.’
In an open letter, CEOs from 46 companies, including Airbus, Lufthansa and Mercedes-Benz, said a delay would signal that Europe is committed to ‘simplification and competitiveness.’