The European Commission is reportedly considering a pause to the entry into application of the Artificial Intelligence Act, citing challenges around technical standards, industry backlash and geopolitical tensions — including direct requests from the U.S. government. At first glance, this looks like prudent regulatory realism. But a deeper look suggests something else: a loss of nerve at the very moment Europe needs genuine leadership. The EU should address implementation challenges head-on — not by deferring ambition, but by doubling down on its digital strategy.
The AI Act is no ordinary piece of legislation. EU policymakers as well as international observers have celebrated the act as the flagship of the EU's claim to global leadership in shaping rules for emerging technologies. From the 2020 white paper "On Artificial Intelligence - A European approach to excellence and trust" onward, its purpose was not just to ensure safety, but to define a European path toward trustworthy, human-centric innovation in AI. A significant delay or full-scale modification of the law, even if framed as technical, would send the opposite message — that the EU no longer has confidence in its own ambitions.
Yes, the AI Act's implementation is difficult. Member states are racing to staff national authorities. The AI Office is still being set up. Harmonized standards still need a lot of work. The Code of Practice for Global Partnership on Artificial Intelligence models remain mired in legal and political complexity, not least due to trans-Atlantic tensions. But none of these challenges justify throwing all of our AI plans into disarray.
In a recently published policy paper, the “European Way. A blueprint for reclaiming our digital future,” my co-authors and I advocated for an EU that moves beyond fragmented policies or erratic actions and instead pursues a coherent values-based digital strategy across the entire technology stack — from e-commerce over data spaces to chips and quantum. The AI Act is one important element of this overall vision. Weakening or significantly delaying it now without a compelling rationale would not only prolong legal uncertainty for our European AI firms, but hollow out one of the EU's most ambitious regulatory projects.
Let us be clear: simplification, clarification and even some recalibration of the AI Act are necessary. Giving the AI ecosystem a few more months to prepare for the application of the new, high-risk obligations might even be imperative. But that's not the same as deregulation. It would be a major strategic error to conflate the need for a sound preparation with a justification for the long-term suspension of the AI Act's enforcement. If anything, it would confirm what many critics have long warned: the EU is strong in drafting legislation but weak in its implementation and enforcement.
Moreover, submitting to geopolitical pressure would be a self-inflicted wound. Washington has reportedly asked Brussels to stop the law's enforcement. The EU should not subordinate its regulatory autonomy to a negotiating chip in trade talks. While the EU must always remain strongly committed to the trans-Atlantic partnership, it must also defend its foundational values. After all, it is precisely the EU's insistence on fundamental rights, risk classification, accountability and transparency that distinguishes its AI approach from laissez-faire or state-driven models.
The real lesson from this whole debate is not so much that the EU has overregulated the tech sector, but that it underestimated what an effective policy strategy really entails. The "European Way" policy paper fills this vacuum by developing a strong overall digital vision, complemented by a coherent policy roadmap. What does it have in store for the AI Act? Instead of retreating from the law, the Commission should use this moment to take three bold steps:
First, it should consolidate the scattered efforts to support small and medium-sized enterprises. SMEs need support that is both authoritative and useful, not just an extensive patchwork of guides, some with questionable correctness, published by numerous public institutions, initiatives and private companies. The Commission could launch a comprehensive easy compliance package for SMEs, bundling legal guidance, flowcharts, checklists, and establish an ambitious EU one-stop shop, using regulatory sandboxes and giving financial incentives. Compliance should be a path to innovation — not a bureaucratic minefield.
Second, it should reconsider whether AI Act enforcement should really be left to decentralized authorities, and whether it should really be separate from the enforcement of other digital regulation. One option would be to move towards a digital enforcement agency with a clear mandate to operationalize the AI Act and related digital laws, such as the EU Digital Services Act or EU General Data Protection Regulation. Decentralized or political enforcement by member states or the Commission risk divergence and regulatory hesitation. Strong central capacity is key to legal certainty and tech investor confidence.
Third, it should roll out a comprehensive "Digital Industrial Strategy" focused on building sovereign digital infrastructure where it matters most: AI compute, AI quality with measurable indicators, resilient cloud, trusted connectivity and applied AI. This is not about replacing global partners, but about ensuring Europe can shape — and not just consume — tomorrow's technologies and make better use of our strengths as well as strategic advantages.
Each of the three steps will face fierce resistance. None will be easy. But then again, neither was the creation of the single market, the launch of the euro, nor the EU's unified response to the COVID-19 or Russia's war against Ukraine. Every milestone in European integration began in doubt and debate — and ended as a defining moment of strength and solidarity. The EU now stands before another such moment. Let us not hesitate. Let us write the next chapter in Europe's story — one defined by purpose, not pressure.
All expressed views are personal and do represent neither the position of the European Parliament nor of the EPP Group.
Kai Zenner is head of Office and Digital Policy Adviser for MEP Axel Voss (European People's Party) in the European Parliament.
Sebastian Hallensleben is chief trust officer at Resaro. He is also co-chair of the AI risk and accountability work at OECD and chair of CEN-CENELEC JTC 2.
