AI and Europe: power shifts, institutions falter (3 of 8)

Peter de ReusMay 1, 2025

Is it a coincidence that Europe is falling behind in the global AI race? AI and Europe are at a tipping point. What was once seen as a technological gadget is now revealing itself as a systemic transformation — one that quietly reshapes our democracy, decision-making, and society.

This is the third blog in an eight-part series in which we explore what AI really is — and why it confronts our political systems with an existential choice. Because behind every algorithm lies a worldview. And behind every European compromise, there’s a clash of national pride, pragmatism, and fear.

Table of Contents

1. AI and Europe: invisible shift of power

AI is not a robot or an app. It is a general-purpose system technology — like electricity or the internet once were — that cuts across all sectors and all facets of our lives. AI recognizes patterns, makes predictions, and generative AI ‘learns’ based on interaction. This means it is increasingly involved in decision-making: about who gets credit, what medical diagnosis is suspected, or what shows up in your news feed. Not based on human judgment, but on statistical correlations — often without us (yet) fully understanding how it reaches conclusions.

This shifts the core of decision-making: from humans to models, from public institutions to partly invisible technical systems, often managed by private companies. And that’s exactly what makes AI so important — not because it’s technology, but because it subtly reshapes power, responsibility, and norms. It doesn’t demand technological fascination but societal vigilance. The table below summarizes the effects of AI.

Effect	Explanation
Decision-making power shifts to systems	AI doesn’t decide independently, but filters what information professionals receive (medical options, credit risks, legal precedents). This shifts influence toward algorithm designers (often tech companies) and away from traditional authorities (governments, doctors).
Policy becomes a data race	Political ideals (e.g., solidarity) compete with statistical logic (e.g., “70% chance of reoffending”). Policy increasingly involves balancing values with predictive optimization.
Global technology vs. local laws	AI systems (think ChatGPT, cloud services) operate across borders, while legislation remains national. This leads to power vacuums: who is accountable if a Dutch hospital uses an American AI diagnostic tool?
‘Human’ work is being rewritten	AI doesn’t just replace routine tasks but also those requiring cognitive creativity (giving advice, editing texts). This pressures jobs (which tasks remain uniquely human?) and income (how do you reward work if AI takes over the “thinking” part?).
Risk of lagging legislation	AI evolves faster than lawmakers can regulate. Think of deepfakes or algorithmic discrimination: the damage is done before rules exist. This creates a constant risk of regulatory paralysis.

The question is therefore not just what AI can do technically, but what kind of Europe it will shape — because every application of AI touches on deeper issues of governance, responsibility, and legitimacy.

2. Can AI and Europe become friends?

AI confronts Europe with a fundamental choice. Not because it requires a standalone technology policy, but because it affects core questions: who decides, how policy works, what work is, and how quickly institutions can adapt. Four visions of Europe — the community of nation-states, the strategic bloc, the federal democracy, and the free market — each offer different answers to these questions, and thus, a different vision of the future.

Vision	Core idea and explanation
Community of nation-states national autonomy Risk: fragmentation, dependence on foreign tech, loss of control.	This vision sees the nation-state as the center of democratic legitimacy and cultural identity. European cooperation is useful, but should be limited to practical matters like trade or security. Sovereignty remains anchored nationally: Brussels facilitates, but does not govern. Explanation: fragmentation weakens Europe’s AI position The vision of Europe as a community of nation-states is based on the idea that sovereignty must remain national. Cooperation is acceptable as long as it doesn’t entail a transfer of power. In the AI era, this approach quickly reaches its limits. AI systems operate across borders, are fed by global data, and run on infrastructure that ignores national frameworks. If every country develops its own rules and standards, fragmentation occurs, Europe’s bargaining position weakens against Big Tech, and dependency on foreign technology grows. National control is maintained in name, but eroded in practice.
Strategic bloc collective resilience Advantage: scale, impact, coordinated response.	This view does not see Europe as a value-based community, but as a geopolitical actor. In a world of global power blocs (US, China), Europe must organize itself as a counterweight. Sovereignty is shared when strategically necessary — in areas like technology, energy, or defense. Explanation: resisting the AI competition The vision of Europe as a strategic bloc emphasizes scale and capability: AI is seen as a geopolitical domain where Europe must be organized to matter among global powers like the US and China. This view enables joint regulation, investments, and market protection, as already seen with the AI Act and Digital Markets Act. It acknowledges AI’s systemic pressure and offers a realistic response: not a full political union, but strategic coordination in crucial areas. The risk lies in legitimacy — if citizens are not involved, technocratic governance without support looms. But if the bloc operates transparently and anchors public values, it can both contain and harness AI.
Federal democracy supranational democracy Promise: democratic legitimacy for large-scale technology.	This vision sees Europe as a political community in the making, based on shared values, rights, and institutions. The EU should evolve into a full-fledged democracy with its own parliamentary sovereignty, supported by European citizens themselves. Explanation: democratic counterbalance in an AI world order The federal democracy takes it a step further: it sees AI not only as a technological challenge but also as a reason to make European governance more democratic and structurally robust. This means organizing power at the EU level and politically legitimizing it through a true parliament and public sphere. This could embed AI policy in values, law, and civic engagement — preventing it from becoming an elite project. At the same time, this model is politically fragile. There is little broad support for a supranational democracy, and institutional reforms are slow. As a result, the vision is normatively strong but limited in short-term feasibility.
Free market economic freedom Consequence: rapid innovation, but risk of monopolies and public vulnerability.	In this view, the EU is not a political entity but an economic platform: an internal market that promotes competition and reduces barriers. Member states retain their political autonomy; the EU primarily ensures that businesses can thrive. Explanation: AI and Europe as a playground for foreign tech giants The free market view sees Europe mainly as an economic space. AI is regarded as a driver of innovation, best served by competition, flexibility, and minimal regulation. This can produce short-term technological growth but overlooks AI’s social impact. Without strong public frameworks, monopolies emerge, labor conditions become precarious, and dependency on foreign platforms grows. In the long run, this vision delivers instability rather than resilience — especially when public values like privacy, justice, and social protection come under pressure.

Europe at a crossroads: organize power or cling to past myths

The rise of AI forces Europe to find a new balance between scale and legitimacy. Those who cling to national frameworks lose control. Those who build large-scale structures without public support lose trust. The vision of Europe as a strategic bloc currently offers the best balance: strong enough to counter global AI powers, and feasible within today’s political reality. But this model will only hold if anchored in democratic processes — with transparency, public engagement, and protection for those impacted by technological change.

AI not only accelerates the policy agenda, it also reveals which European mindset still works. The choice is clear: either Europe organizes itself strategically, or it is sidelined — with growing loss of autonomy, influence, and prosperity as a result.

3. Europe governs, but barely listens

The EU often governs in a top-down manner: policy is crafted in Brussels, rolled out across 27 countries, and rarely adjusted when it clashes with real-world conditions. What is intended as common policy often feels like a straitjacket. Citizens and sectors negatively affected by the rules rarely get a hearing, because the system is slow and barely self-correcting.

Behind the ideal of a stronger Europe often lies a centralist model, modeled after a country like France. This breeds resentment: cooperation is experienced as imposition. And when policy has a disproportionately heavy impact in a specific country or region — such as nitrogen regulations in the Netherlands — there’s no mechanism in place to correct it.

The problem is fundamental: open borders, housing ambitions, and strict environmental goals are difficult to reconcile at the national level. We cannot simultaneously allow unrestricted migration, build housing on a large scale, and reduce nitrogen emissions in the same territory — especially not when neighboring countries apply much looser standards. For many citizens, this feels like unfair policy imposed from above and beyond their influence. Powerlessness then quickly turns into resentment, not just toward The Hague, but also toward Brussels.

4. AI and Europe as a mirror of division

Europe also deserves recognition. In Brussels, many competent and thoughtful people work on legislation that is often more robust than what individual member states have ever produced. Consider initiatives like the AI Act or the Digital Markets Act — efforts to provide direction and counterbalance in a global tech landscape. The intent is usually rational, orderly, and aimed at the common good. Yet there’s sometimes a deep gap between policy logic and societal reality.

National myths as obstacles

What complicates Europe is that each country clings to its own myths and rarely takes a critical, analytical look at itself. National pride, deep-rooted rituals, and collective beliefs shape how policy is made and experienced — even when those beliefs diverge from facts.

During a boat trip in Greece, an old man said to me: “Do you know what the biggest bank in the world is? The National Bank of Greece.” Throughout the trip, Greece was number one. Every country has its blind spots.

It would be unfair to single out just one country. Every member state has a collective self-image that often only partially aligns with reality. The Netherlands sees its legal system as a beacon of reason, but in practice it is erratic, unpredictable, and more biased than we care to admit. Germany prides itself on moral seriousness but clings to an almost dogmatic belief in privacy — even when it comes at the expense of innovation or openness. France views rules as the foundation of civilization, but the moment someone questions their effectiveness, those same rules are suddenly sacrosanct. And Spain tends to keep foreign expertise and help at arm’s length out of pride — yet is first in line when Brussels distributes billions.

AI as a tool against complacency

Every country — and even regions within countries — clings to its own identity, rooted in supposed tradition. But those traditions are often intertwined with commercial interests, social positions, and symbols. Among the national elite, the saying often applies: better the head of a mouse than the tail of a lion. But who are they really serving — the citizen, or themselves?

And shouldn’t journalists look beyond power and their loyal, aging base? Precisely where things get uncomfortable. Shouldn’t there be more focus on the youth — not to lecture them, but to acknowledge that they have different interests and ask different questions?

It’s precisely in this struggle that AI can be a powerful tool — not as a substitute for thinking, but as a mirror for facts, perspectives, and our own assumptions.

5. AI and Europe: Weaknesses exposed through in-depth AI analysis

Generative AI helps us uncover weaknesses — not with images, visions, or opinions, but with data and benchmarking. AI draws connections, finds exceptions, identifies structural flaws. Not to replace political judgment, but to provoke sharper questions. And that’s exactly what Europe needs: not a new blueprint, but the courage to reflect, learn from one another, and improve together.

If news media and journalists want to remain relevant, a drastic shift is needed: AI must be used as a journalistic tool. Because what generative AI reveals cannot be put back in the bottle.

Below are some examples illustrating this approach — not as conclusions, but as invitations to further analysis and discussion.

Netherlands: judicial interpretive freedom examined

Dutch judges have a notably wide margin of interpretation. This AI-assisted investigation analyzed family law and contract law in particular. As a comparison, Germany, France, and Spain were also examined — countries where judges are generally bound by stricter standards, for example in how they handle statutes and fact-finding.

The findings were translated into three short videos, based on the classic model of thesis, antithesis, and synthesis:

Thesis
(the claim, in German)

Antithesis
(the defense, in Dutch)

Synthesis
(the connection, in Dutch)

The videos are based on two underlying publications:

A comprehensive comparative AI study on interpretive freedom in four European legal systems.
An accessible classroom version for Dutch students in VMBO-4 and VWO-6.

France: capital gains tax on real estate

The plus-value immobilière was historically introduced in France as a tool for tax fairness: those who profit from real estate must contribute a share to society. The rationale behind it:

Profit from property is a form of passive income and should be taxed like labor.
Property value often increases due to collective investments (in infrastructure, safety, economy), not just individual effort.
The tax aims to discourage speculation and promote long-term, responsible ownership.

This tax is deeply rooted in the French tradition of égalité and social correction, taxing wealth that passively increases in value to counter inequality. However, this belief is now at odds with new housing market realities.

1. Situation: fiscal incentive vs. current market dynamics
France levies a capital gains tax of up to 36.2% on sales of second homes. Exemption applies only after 22 years (income tax) and 30 years (social contributions).

Facts:

3.1 million homes are vacant (8.2% of housing stock)
(source)
In the Netherlands, that’s just 3.4% (~279,000 homes); in Germany 4.4%
(Eurostat)
Revenue from the plus-value is relatively small: €1.8 billion per year
(source)

2. Problem: morally legitimate tax, economically inefficient

Owners delay selling, clogging the market
Vacancy rises, especially in tourist regions and inherited properties
(Le Monde)
Mobility and reinvestment are stifled
Government misses out on revenue from other sources

3. Impact: structural harm to housing market and state revenue

Fewer transactions → lower transfer taxes and VAT
Real estate becomes a store of value, not a housing solution
Aggravation of the housing shortage
Reduced economic activity (renovations, relocations)

4. Change: eliminate the tax
Propose a reform abolishing the capital gains tax.

5. Outcome: activate housing, strengthen the state
By reducing vacancy to Dutch levels:

Increase housing stock without new construction: 1.8 million homes reactivated
Higher revenue through transactions: potential transfer tax (5.81%): €27.6 billion
Improved market dynamics
Maintain tax fairness through other instruments

Germany between principle and practice: privacy as a brake on digitization

The German right to informational self-determination is deeply rooted in the constitution and collective memory — a response to surveillance excesses during the Nazi era and the GDR.

The state should not know everything.
Individuals must retain control over their data.
Privacy is not just a right, but a safeguard against state power.

These principles form the moral backbone of German digital policy — but increasingly clash with economic and social realities.

1. Situation: legally strong, digitally weak

Germany invests €3.3 billion annually in government digitization.
Only 46% of citizens use e-government services (EU average: 70%, Netherlands: 96%).
In 2024, public services still rely on fax machines, paper forms, and physical files.
eID usage is just 9% (vs. 90%+ in NL).

Despite high spending, usage and impact lag far behind comparable countries.

2. Problem: principled privacy hinders digital progress

Strict legislation and extremely cautious data policies
Federal fragmentation: 16 states with their own IT policies and privacy commissioners
Resistance to automation in government, healthcare, and mobility

Compared to the Netherlands:

Netherlands spends less overall (€1.2 billion), but more per capita: €69 per person vs. €39 in Germany
(CBS).
NL reaches 96% of its population with e-government services, vs. only 46% in Germany
(source).
Citizens and businesses in the Netherlands benefit more quickly from digital transactions, permits, tax services, etc.

3. Impact: tens of billions lost in annual growth

Factor	Germany	Netherlands
Use of e-government services	46%	96%
GDP (2024)	€4,450 billion	€1,070 billion
Growth lost due to digital lag	Up to €111 billion/year	Negligible

Each 1% lower usage of e-government = approx. 0.05% less economic growth
Germany misses out on 1–2.5% GDP growth/year (€44–€111 billion)

4. Change: from privacy paranoia to public data utility
Proposal:

Create a national data framework for public value, where:

Anonymized data can be used for healthcare, mobility, and energy.
Faster approval processes for research data.
Responsible data policies overseen by an ethics council, not just lawyers.

5. Outcome: preserve trust, restore growth

Annual economic gain: €44–€111 billion
Stronger public services
More room for AI, innovation, and startups
Maintain public trust through a transparent, ethical framework

This way, Germany can remain Europe’s privacy champion without sidelining itself economically.

6. AI as a mirror for Europe

Whether it’s French regulations that block the housing market, or German principles that hinder digital progress — AI makes painfully clear where policy gets stuck in entrenched beliefs. Not because those beliefs are nonsense, but because they are rarely tested against a changing reality. AI can do that: by seeing connections humans can no longer spot, by distinguishing myth from impact.

But only if we dare to use it as a mirror. Not to undermine national pride, but to make policy smarter, fairer, and more future-proof. Then AI becomes not a threat to our identity, but an invitation to reshape both AI and Europe — not as a top-down blueprint, but as a collaboration that actually works — precisely because it dares to observe, learn, and adjust.