Tech entrepreneur and investor Xavier Buck shares a founder’s perspective on Europe’s struggle to balance innovation with regulation. With over 25 years building internet and AI companies, he offers a candid take on the EU AI Act—its promise, its pitfalls, and what it means for Europe’s creative sovereignty.
After 25 years of building internet and AI-driven companies across Europe, I still see the same pattern repeat itself. Brussels regulates first and builds later. Every technological wave; search, social, blockchain, and now artificial intelligence. It begins with excitement and ends tangled in administration and red tape.
The AI Act Paradox
The EU AI Act is a historic attempt to make algorithms safe while keeping innovation alive. It classifies systems by risk: bans for “unacceptable” uses, strict checks for “high-risk” ones, transparency for “limited-risk,” and freedom for “minimal-risk.” It also extends to foundation models and general-purpose AI. The ambition is noble and the logic sound, but the Act risks suffocating the very builders it seeks to empower.
“The ambition is noble and the logic sound, but the Act risks suffocating the very builders it seeks to empower”
To be fair, this time Europe provides some predictability. Founders can place themselves within a clear legal framework and plan accordingly. The introduction of regulatory “sandboxes,” controlled environments where innovators can test alongside regulators, is encouraging. Trust, transparency, and fairness can become real advantages if applied with pragmatism.
The Founder’s Red Flags
Yet the old problems return. Compliance costs crush small teams. Twenty-seven national authorities will mean twenty-seven interpretations. Lawyers come before engineers, caution before creativity. When legal risk outweighs reward, talent quietly leaves for places where experimentation is celebrated instead of punished.
The Act’s demand to disclose every dataset used for training misunderstands learning itself. Models do not copy; they learn patterns. If we treat learning as plagiarism, we are not protecting society, we are freezing intelligence.
The One Billion Euro Mirage
In October 2025, Brussels proudly announced “one billion euros for AI.” It is being deployed through Horizon Europe and Digital Europe, the same large frameworks that were meant to support innovation for years. Yet anyone who has applied knows how these programmes work. Most of the money ends up funding studies, coordination groups, and consultants, while truly impactful projects rarely see support. We have seen countless calls, evaluations, and reports, but very few real outcomes. Europe keeps missing the right projects because its process rewards paperwork, not performance.
“Most of the money ends up funding studies, coordination groups, and consultants, while truly impactful projects rarely see support”
Bureaucracy Over Brilliance
I have lived this frustration firsthand. Our consortium of three specialised European companies submitted the best technical design for a sovereign DNS-resolver project: secure, resilient, and fully European. But we applied under a single consolidated European entity instead of three separate legal ones, and that procedural detail alone disqualified us. We were not penalised for weak engineering but for failing to follow a bureaucratic ritual that only insiders know. Evaluators rewarded form over substance.
That is how Europe keeps losing, not through lack of talent but through addiction to process. The system punishes those who build and rewards those who file.
Regulation Without Enablement
The EU calls itself a techno-democracy, but innovation is meritocratic. Success depends on execution, not reports. While Brussels debates “gender balance in work packages,” others release working products.
True enablement would be simple:
- One European Startup License with one registration, one tax ID, and one payroll system.
- A single funding portal run by people who have actually built companies.
- Decisions within 60 days, not 16 months.
- Minimum Ten percent of Big Tech fines reinvested into a Sovereign Tech Fund.
- Priority public contracts for startups that comply with the AI Act.
“Founders are not against rules; we are against rules without support”
Founders are not against rules; we are against rules without support. When policymakers focus only on removing risk, they remove ambition too. Europe becomes the safest place to start a company and the hardest to scale one.
Creative Sovereignty
If Brussels truly wants digital sovereignty, it must learn to trust its builders. The AI Act could still evolve into a partnership instead of a prison if implemented with humility and flexibility.
Europe does not need fewer rules. It needs rules that learn, adapt, and improve with time.
Because true European sovereignty is not regulatory.
It is creative.
This article was published in the Silicon Luxembourg magazine.
