We are irrelevant already. Only German tanks “Leopard” will save us. We will use them to herd goats in the bushes. We will have milk again. This is the basis of survival of our continent in tough times, as history shows. Thank you, Mr. Merz.
“The European Union requires a strategic emergency plan for its economic and digital survival. There is currently no such thing as digital self-sufficiency.
The USA dominates artificial intelligence (AI), cloud computing, and digital platforms.
China is building its own technosphere with global reach.
So far, Europe has failed to provide a third model—offering only regulation without power, visions without investment, and coordination without resolve. It lacks strategy, courage, creativity, political leadership, and decisive action.
The EU focuses on regulation, losing time, talent, and technological substance in the process. Attempts to reduce bureaucracy have been a joke—too little, too late. If democracy is to survive, its sluggish structures must be adapted—despite political fragmentation—and quickly, as time has become a critical factor for survival.
This is a diagnosis of failure—structural, political, and institutional.
Digital sovereignty without a sovereign infrastructure is an illusion.
What we currently lack is the absolute will for digital—and therefore economic—survival.
The EU must reinvent itself strategically, economically, socially, and politically. To successfully navigate these transformations (climate, digitalization, energy, defense), we need concrete legislative action; anything else is mere wishful thinking.
The decisive battleground is AI. Whoever controls it, controls the economy, the military, and culture.
Europe possesses excellent research, yet it lacks a single competitive ‘frontier’ model. Not enough compute.
The AI Act creates legal uncertainty rather than competitive advantage, the European AI action plan lacks both enforcement power and funding.
A CERN-style model for AI—public-private, science-based, and open—would be a good starting point, complemented by a "Eurostack" as a sovereign digital infrastructure. Europe should invest in "Scientist AI"—systems trained on truth rather than on what is merely pleasing—which would be both more powerful and more secure.
Whoever builds the more secure AI will possess a more credible product for export.
What needs to be done now—seven immediate decisions:
1. Complete the EU Digital Single Market: We need unified rules rather than 27 different national sets, along with a three-month "fast-track" mechanism for urgent digital legislation.
2. Create a genuine Capital Markets Union: This requires more venture capital, a €50 billion European scale-up fund for late-stage financing, and incentives to attract talent. Private capital, —including funds from pension schemes and insurance companies—must be leveraged through attractive programs offered by the European Investment Bank and the European Investment Fund. Currently, both institutions are far too slow, risk-averse, and bureaucratic to keep pace with the dynamics of the digital economy.
3. Centralize digital oversight: Europe needs a single, unified authority for digital decision-making and enforcement, replacing the current 27 national bodies.
4. Conduct an audit of dependencies, complete with mandatory exit strategies for every critical technological dependency.
5. Implement innovation-friendly procurement policies, reserving 20 percent of public IT spending for young European start-ups.
6. Establish "digital special economic zones": Currently, it takes European start-ups three to five years to ensure compliance with regulations across the entire EU—a form of structural self-sabotage. Within these designated legal zones, digital companies should be permitted to operate under radically simplified rules for up to seven years—creating "scaling spaces" that allow them to compete effectively with global rivals.
7. Cybersecurity and the Defense of Democracy: Hybrid warfare against Europe is a present-day reality—characterized by disinformation, election interference, and troll farms designed to divide us. The goal is paralysis: an EU so preoccupied with its own internal affairs that it remains immobilized. To save democracy, we need bans on bots, oversight of algorithms, and a robust Digital Services Act (DSA)—simply verifying the age of young users does nothing to address the content itself.
Quantum computing, robotics, biotechnology, and autonomous systems will be the battlegrounds of tomorrow.
Within five years at the latest, network effects will effectively close off markets; any entity that fails to invest massively and strategically by 2026 will find itself a "digital colony" by 2031.
Europe’s strategy must be one of "leapfrogging"—taking the lead where industrial strength and scientific excellence converge, rather than trying to catch up in areas already dominated by others.
This includes implementing European "tech visas," making computer science a mandatory subject starting in the first grade, and creating interoperable data spaces that finally transform Europe’s data capital [A] into the foundation for its own AI.
Alongside a space-based digital system, a framework for "digital diplomacy" must be institutionalized at the European level immediately.
We can no longer afford 27 separate national digital strategies or endless consultation processes. If every member state does not pull in the same direction, a Europe of varying speeds is unavoidable. The alternative to such a radical shift is irrelevance. The decision rests with the heads of state and government—and it must be made now.
Axel Voss is a CDU member of the European Parliament.
The EU requires a strategic emergency plan for its economic and digital survival—yet, thus far, the absolute determination to act is missing.” [B]
A. Does Europe’s data capital as basis for building AI exist? Who will give away their trade secrets to create it?
Europe’s "data capital" exists as a vast reservoir of raw industrial, healthcare, and public-sector information, but it does not yet exist as a unified, ready-to-use pool for training AI.
Roughly 80% of European industrial data remains trapped in private corporate silos.
Furthermore, companies do not actually have to "give away" their trade secrets to unlock it—the European Union is instead forcing and incentivizing data access through strict laws and privacy-preserving architecture rather than voluntary charity.
Does Europe’s AI Data Capital Exist?
Europe holds massive amounts of valuable non-personal data—particularly in manufacturing, automotive, energy, and healthcare—but faces a major gap between having data and mobilizing it.
• Fragmented Silos. Unlike the consumer internet data scraped by U.S. tech giants, Europe's strength lies in proprietary industrial and institutional records. Because this data is scattered across thousands of companies and national borders with different standards, AI developers cannot easily access or aggregate it.
• Federated Infrastructure. Instead of building one giant centralized database in Brussels, the EU is building Common European Data Spaces and AI Factories.
Networks like Catena-X (for automotive supply chains) and EUCAIM (for cancer imaging) link separate databases together so AI models can train on distributed data without moving it to a single owner.
• Compute Bottlenecks. Even where data is available, Europe accounts for only about 5% of global AI computing capacity.
The EU is investing billions into AI gigafactories to give startups and researchers the hardware needed to actually process this data locally.
Who Will Give Away Their Trade Secrets?
No company willingly surrenders its core intellectual property to competitors. To solve this standoff, the EU relies on legal mandates, technical safeguards, and commercial trade-offs rather than asking firms to give away secrets for free.
• Mandatory Sharing Under the Data Act. Under the EU Data Act, manufacturers of connected products—such as smart industrial machinery, medical devices, and vehicles—are legally required to share raw and pre-processed user-generated data with users and authorized third parties. Companies cannot simply invoke blanket "trade secret" excuses to block access to this operational data.
• "Protect, Then Share" Safeguards. While companies must share operational data, the law enforces a "protect, then share" rule. Data holders can require strict non-disclosure agreements, ban reverse-engineering, and use Secure Processing Environments where an AI model trains on encrypted or isolated data without the developer ever seeing the underlying raw files. If sharing would cause serious, irreparable economic harm, companies can pull a legal "trade secrets handbrake" to refuse disclosure.
• Commercial Incentives and Sovereignty. Many European firms refuse to upload internal data to U.S. cloud providers out of fear that American tech giants will absorb their corporate know-how. By participating in European data spaces, companies get paid compensation for licensing non-core data while gaining shared access to industry-wide AI models—such as predictive maintenance tools—that no single company could build alone.
B. Europa digital oder irrelevant. Frankfurter Allgemeine Zeitung; Frankfurt. 14 Aug 2026: 18. Von Axel Voss
Komentarų nėra:
Rašyti komentarą