When Tesla announced its first European factory near Berlin in late 2019, the state of Brandenburg treated it as a national priority. Every available fast-track mechanism was deployed. And yet the process still required 19 separate preliminary building permits. The factory was physically complete – walls up, equipment installed, ready to produce cars – before the final construction permit was issued in March 2022. Germany treated the 861 days from announcement to opening as impressively fast.

In China, the legal deadline for issuing a construction permit after a complete application is seven days.

Nobody who has worked inside a European government – and I have spent years inside Lithuania’s – can in good faith dispute the problem. The regulatory burden is real, structural, and costly. The Ifo Institute puts the annual drag on Germany alone at roughly €150 billion in lost economic output. These are not abstract figures from business lobbies; they are the lived experience of companies, developers, and citizens who have learned to budget years, not months, for processes that in other parts of the world take weeks.

At Davos in January, Germany’s Chancellor Friedrich Merz put a sharp point on this reality for the continent: the EU is losing ground to the United States and China, and bureaucracy is a significant reason why. Merz’s prescription was an “emergency brake on bureaucracy.” His diagnosis was largely correct, here’s the thing that tends to get lost in the political excitement of the moment: we have been here before.

Regulatory simplification has been the promise of European governments for at least two decades. The UK tried it under David Cameron with the “one in, two out” rule. It started loudly and quietly failed. The European Commission launched waves of simplification initiatives. The stack kept growing. Last year, Lithuania’s Ministry of Economy and Innovation reported that national legislative efforts had reduced the regulatory burden on business by €21 million – a solid, hard-won result. In the same period, new EU regulatory requirements increased it by €28 million. Net outcome: minus seven. This is not a Lithuanian failure. This is simply what the system does. The regulatory environment is not a fixed object you can whittle down with enough political will. It is a living one. It regenerates.

This matters enormously, because it changes the diagnosis – and therefore the cure.

The seven-day Chinese comparison is not entirely fair, and it should not be taken as an argument that German standards are too high. China’s speed comes at a cost Europe has deliberately chosen not to pay: weaker public participation, fewer environmental safeguards, less protection for the people and communities that regulations exist to serve. The lesson of the Tesla factory is not that German rules are wrong; it is that even when every political actor wants the same outcome, the administrative machinery cannot deliver it at anything close to the speed the situation demands.

The bottleneck, in most European permitting and compliance processes, is not the standard itself; it is the processing around it. Reviewers reading, cross-referencing, flagging, and coordinating – one application at a time, in sequence. Applicants cycling through rejections for missing documents or minor inconsistencies. Sign-offs chased across agencies that do not share systems. The standard itself often takes an afternoon to apply. Everything else takes months.

None of this means consultation windows disappear. Those have their own clocks and exist for substantive reasons. The point is that in most European permitting processes, administrative latency – reviewers reading, applicants cycling through rejections, signoffs chased across agencies – is the larger share of elapsed time, and it is the part AI compresses. The Tesla factory was not held up by an 861-day public comment period. It was held by paperwork.

The political appetite for mass elimination of regulation simply does not exist in Europe, nor should we expect it to. The goal, therefore, cannot be to strip away the regulatory architecture. The goal must be to traverse it faster.

This is where the conversation needs to shift – sharply and urgently – because the technology to do exactly this already exists.

AI As The Accelerator

AI systems can work with applicants before a submission is even filed, checking a planning application or compliance document against every relevant local and EU requirement, identifying missing materials, flagging inconsistencies in technical drawings – all before the file ever reaches a human inbox. On the institutional side, the same category of system can process in hours what an overworked reviewer takes weeks to do, then escalate only the cases that genuinely require human judgment. The rules remain identical. The wait does not.

This is not theoretical. Singapore’s CORENET X platform, mandatory for all new building projects from October 2026, pairs applicant-side validators that ensure compliant submissions with agency-side automated checking that consolidates approvals across five regulators – building authority, planning authority, environmental agency, transport, and water – within twenty working days. Same building codes, same standards, different speed.

While most European digital ministries have AI procurements underway, pilots in flight, automation tools being bolted onto existing workflows, the question is whether any of it goes far enough. Singapore did not get to twenty days by adding a chatbot to its permit portal. It rebuilt the submission and review architecture around the assumption that machines would do most of the reading – structured data, codified rules, consolidated approval. That is the bar. Most European deployments are not yet near it.

These tools are not waiting to be invented. They are waiting to be deployed by governments willing to use them.

Let me be direct about what this means for the current political conversation. Deregulation, pursued with the energy it is currently receiving, risks becoming a costly distraction. A government can spend years fighting to strip away regulations, encountering institutional resistance, public backlash, and EU-level headwinds, and arrive – after all of that – at a somewhat lighter rulebook that still takes the same number of months to navigate because the processing infrastructure has not changed. A government that chose instead to instrument its regulatory delivery with AI, without touching a single line of legislation, could plausibly deliver timelines three to five times shorter within a single parliamentary term – the kind of range Singapore is now mandating for its construction permits.

The longer-term work of streamlining and rationalizing the regulatory stack remains important and should continue. But it must not be allowed to become an excuse for inaction on what can be done right now. The technology exists today. The political risk is low compared to the bruising work of legislative reform. And the benefits would be felt immediately – by the businesses applying for permits, by the agencies processing them, and by the citizens waiting for decisions that affect their lives.

A Different Kind of Emergency Brake

Merz is right that Europe has a speed problem, and right that AI is part of the answer. But Europe does not need to choose between its values and its competitiveness. It does not need to race China to the bottom on environmental or labor protections or hollow out the consultation processes that give democratic legitimacy to development decisions.

The emergency brake Merz called for is necessary, but pulling the right lever matters. A continent that spends the next five years fighting to rewrite its rulebook, while leaving its processing infrastructure untouched, will have wasted the best opportunity it has had in a generation. The instrument is in our hands. It is time to use it.

Rimantas Zylius served as Minister of Economy of Lithuania.