EU Considers Delaying AI Act Rollout Amid US and Big Tech Pressure

After months of pressure from American tech giants and the Trump administration, Brussels officials are now discussing potential delays to parts of the AI Act, marking a dramatic shift from their earlier stance.

The Simplification Package

On November 19, 2025, the European Commission will decide on a “simplification package” that could reshape how the AI Act rolls out. According to Commission spokesperson Thomas Regnier, officials are exploring “targeted implementation delays,” though no final decision has been made yet.

The draft proposal includes significant changes. Companies using high-risk AI systems could get a one-year grace period before enforcement begins. The EU is also considering pushing back fines for transparency violations until August 2027. This would give businesses more time to adjust their operations without facing immediate penalties.

The AI Act originally entered force in August 2024, but its rules apply gradually. Some restrictions on dangerous AI systems already took effect in February 2025. The next major deadline is August 2, 2025, when general-purpose AI models must comply with new requirements. The toughest rules for high-risk systems were scheduled for August 2026.

Big Tech Pushes Back

American technology companies have been fighting the AI Act for months. In June 2025, the Computer and Communications Industry Association, which represents Apple, Google, and Meta, asked the EU to delay the law by at least two years. They argued that the rules are too complex and uncertain.

The lobbying campaign has been massive. Recent research shows tech companies now spend about €151 million per year lobbying in Brussels—a 50% increase from four years ago. Meta alone spends €10 million annually, while Microsoft and Apple each spend €7 million.

Meta has taken the hardest line against the regulations. The company’s global affairs chief publicly announced Meta would not sign the EU’s voluntary AI Code of Practice, claiming it creates too much legal uncertainty. In contrast, Google decided to endorse the code, showing divisions within the tech industry.

The US government has also applied pressure. Vice President JD Vance warned at a Paris summit that “excessive regulation” could damage Europe’s AI industry. Behind closed doors, Brussels has been negotiating with the Trump administration about adjusting the AI Act and other digital rules.

A Stunning Reversal

What makes this situation remarkable is how quickly the European Commission changed its position. In July 2025, Commission spokesperson Thomas Regnier firmly stated: “Let me be as clear as possible, there is no stop the clock. There is no grace period. There is no pause.”

That was just four months ago. Now, the same officials are openly discussing delays. This reversal came after more than 45 European companies, including Airbus and Mistral AI, also called for postponing the law.

The change reflects real implementation problems. Key guidance documents arrived late or not at all. The Code of Practice for general-purpose AI was supposed to come out in May 2025 but got delayed due to industry complaints. Some guidance on prohibited AI systems was published just two days after those rules took effect, giving companies almost no time to prepare.

Draghi’s Influential Voice

Former Italian Prime Minister Mario Draghi added weight to the delay arguments in September 2025. Speaking at a Brussels conference, Draghi called the AI Act “a source of uncertainty” and urged pausing the next phase of implementation.

“The first rules—which included the ban on ‘unacceptable-risk’ systems—landed without major complications,” Draghi explained. “But the next stage—covering high-risk AI systems in areas like critical infrastructure and health—must be proportionate and support innovation and development. In my view, implementation of this stage should be paused until we better understand the drawbacks.”

Draghi’s 2024 competitiveness report had already pushed the EU to rethink some regulations. His recommendation to delay corporate sustainability rules was accepted, and now his AI Act concerns are getting serious attention.

Poland has also been pushing for delays. The Polish government proposed a “stop the clock” mechanism for parts of the AI Act where technical standards aren’t ready yet. In September, Poland suggested delaying penalties for high-risk AI systems by six months to a year.

The Opposition

Not everyone supports delaying the law. More than 50 organizations, including digital rights groups and consumer advocates, signed a letter opposing any postponement. They worry that delays will weaken protections for citizens and give too much power to tech companies.

Legal experts have mixed views. Some argue that simply delaying a flawed law won’t fix its problems. Thibault Schrepel, an associate professor at Vrije Universiteit Amsterdam, wrote that the AI Act’s rigid framework can’t keep up with fast-changing technology. Instead of delays, he suggests building “adaptive regulation” that can evolve alongside AI development.

The blockchain and crypto sector is watching these developments closely, as AI regulations could affect projects combining artificial intelligence with blockchain technology.

Balancing Act Ahead

The November 19 decision will test whether Europe can balance innovation with regulation. Even if the Commission proposes delays, the changes would still need approval from EU member countries and the European Parliament.

The stakes extend beyond technology policy. The AI Act was supposed to establish Europe as the global leader in responsible AI governance. Backing down under American pressure could undermine that position and raise questions about EU sovereignty in the digital age.

Companies operating in Europe face their own challenges. They need clear rules to make business plans, but they also want those rules to be workable. Tech leaders have complained that products get delayed or watered down because of regulatory uncertainty.

The bigger question is whether any law can effectively regulate technology that changes so quickly. General-purpose AI models have evolved faster than regulators expected. What seemed like comprehensive rules in 2021 may already be outdated in 2025.

Reality Check

Europe now faces a difficult choice on AI regulation. The November 19 package could either provide breathing room for better implementation or signal retreat under corporate pressure. Whatever happens, the decision will show whether democratic governments can set technology rules or if tech giants hold too much power. The world is watching to see if Europe’s ambitious AI law survives contact with reality.