In contrast, the EU’s newly announced “AI Continent” Action Plan feels cautious and incremental. If Europe genuinely intends to compete in AI, it must be bold—doubling down on open-source innovation, increasing R&D investment, and integrating AI into its defense strategy. Crucially, it should seek collaboration with US tech leaders rather than shutting them out.
Here’s a breakdown of the EU’s five-pillar strategy—followed by realistic, more ambitious alternatives.
1. Infrastructure
The EU plans to invest €2 billion in academic supercomputers and €20 billion in five AI gigafactories, with the aim to triple data center capacity by 2032. This is modest compared to trillion-dollar initiatives in the US and China.
If the EU were serious:
-
Shift gigafactory funds to cloud infrastructure. Support commercial providers through subsidies and long-term contracts to accelerate the data center target to 2028.
-
Partner with US hyperscalers. Offer full market access in exchange for data localization, EU-based R&D, and compliance with European security rules—transforming reliance into local capability.
2. Data
While data fuels AI, Europe’s strict privacy laws and fragmented regulations stifle access. Meanwhile, the US and China benefit from vast, flexible data ecosystems.
If the EU were serious:
-
Modernize GDPR. Allow safe harbors for anonymized, high-value datasets and streamline data processing under “legitimate interest.”
-
Enforce and expand open data. Push governments to go beyond minimum requirements and maintain a centralized, searchable dataset registry.
-
Incentivize data sharing. Provide grants or tax breaks to entities releasing high-quality, open-license datasets.
3. Adoption
Only 13% of European businesses consider AI central to their operations—far behind the US (49%) and China (83%). Brussels’ traditional playbook—white papers, pilot programs, and innovation hubs—is too slow.
If the EU were serious:
-
Launch bold, measurable AI missions. Examples: 90% of cancer screenings read by AI by 2030; 25% of traffic lights adaptive by 2027; real-time crop monitoring on 50% of farms by 2030.
-
Procure results, not paperwork. Redesign public contracts around outcomes (e.g., “cut ER wait times by 20%”) and open competition to AI-native startups.
4. Skills and Talent
The EU’s AI Skills Academy and minor migration reforms won’t reverse the ongoing brain drain to the US. Silicon Valley already attracts a disproportionate share of Europe’s top talent.
If the EU were serious:
-
Create 1,000 AI Chairs. Ten-year, well-funded positions to attract global AI leaders back to Europe.
-
Redirect Erasmus and Social Funds. Allocate 10–20% to AI-heavy joint degrees and specific tracks like AI & medicine or law. Dedicate a portion of the €140B social budget to AI reskilling.
-
Track talent. Publish a continent-wide AI Talent Scoreboard with data on degrees, salaries, and migration.
5. Regulation
The EU has signaled flexibility on AI regulation but has delivered little concrete simplification. The proposed AI Act Service Desk offers guidance, but doesn’t change the rules themselves.
If the EU were serious:
-
Streamline B2B compliance. Create safe zones for low-risk use cases and fast-track approval for high-risk ones via pre-certified templates.
-
Enable EU-wide “passporting.” One national approval should grant access across the Single Market—something surprisingly absent in the current AI Act.
-
Exempt mature, low-risk applications. Tasks like public transport routing or manufacturing inspections should be automatically approved.
What’s Missing?
Notably, the EU plan is silent on:
-
Defense and Security. Ukraine’s battlefield success has shown the power of AI-enabled warfare. Europe should fund military AI through an outcome-driven model, akin to the US’s DARPA.
-
Open Source. Open-source AI was 2023’s unexpected disruptor. Europe should lean into this model by expanding funding and prioritizing openness.
-
Global Alliances. As the US looks inward and China’s governance remains problematic for democracies, Europe has a chance to lead a “rest of the world” alliance—working with the UK, Japan, South Korea, Canada, Ukraine, and others to create an open, rules-based AI ecosystem.
While opinions may vary on the specific actions proposed, the need for bold, decisive steps is not up for debate. Europe has the talent, capital, and market to compete—but only if it acts with urgency and ambition.
Luukas Ilves is a non-resident fellow at the International Center for Defence Studies in Tallinn and advisor to Ukraine’s Vice Prime Minister. He previously served as Estonia’s Chief Information Officer.