In Brussels this week, the European Commission unveiled RAISE — the Resource for AI Science in Europe, a new pan-EU infrastructure intended to accelerate research and development in artificial intelligence.
It will be backed by €107 million in Horizon Europe funds, and aims to federate data, compute resources, and expertise.
It is a bold gambit. But in the era of AI rivalry between the United States and China, Europe’s strategic position is far from secure. The launch of RAISE presents both an opportunity and a challenge — one that illuminates the broader contradictions in Europe’s approach to technological sovereignty.
Europe’s AI ambition: credible or cosmetic?
On paper, RAISE is precisely the kind of strategic infrastructure Europe has lacked. Federated compute, shared data sets, harmonised standards — these are the building blocks of a competitive AI ecosystem. Brussels hopes RAISE will reduce duplication between member states, foster cooperation, and help smaller nations plug into a continental core.
Yet the scale is telling: €107 million, though a respectable sum in research funding, is modest in comparison to the scale of AI investments in Wall Street labs or Chinese state-backed initiatives. For example, private U.S. AI firms routinely attract billions in venture capital — each. Europe’s commitment must be judged not only by its gesture, but by its persistence and ability to draw private inflows.
Moreover, expertise remains fragmented across the bloc. Too many universities, too many national programmes, too few centres of excellence that punch above their weight globally. RAISE may knit together offerings, but real competitive weight comes from a handful of ‘magnet’ institutions — which require focus, critical mass, and a culture of risk.
The sovereignty illusion
European technocrats love the language of “strategic autonomy.” But autonomy in AI is a mirage unless Europe can produce, deploy, and maintain its own innovations at scale. Infrastructure like RAISE helps, but it cannot substitute for the ecosystems of capital, regulation and market incentives dominated by Silicon Valley and Shenzhen.
Brussels must reconcile two impulses: on the one hand, control — stringent data protection, algorithmic regulation, liability laws — and on the other, dynamism — freedom to experiment, fast commercialisation, risk tolerance. The temptation is to overregulate and stifle, particularly under the weight of privacy and caution. RAISE must resist being mired in red tape if it is to serve as a springboard, not a straitjacket.
Even if RAISE becomes operational, its success depends on member states buying in. That means aligning national digital strategies, pooling sovereign compute infrastructure, and ceding a degree of control. Europe has long struggled with this: subsidies, prize funds, regional winners and losers, national ego clashes.
If Germany, France, the Netherlands or Italy divert resources to their own national AI hubs instead of fully embracing the shared infrastructure, RAISE may become a “shell” — present in name, hollow in effect. Worse, wealthier states could cherry-pick — grabbing the lion’s share of capacity, leaving others marginalized.
Brussels must insist on equitable governance, transparent allocation of resources, and incentives for weaker states to join wholeheartedly. Political unity will be the acid test of the project’s ambition.
Industrial backing and the private sector stake
Here lies the make-or-break factor. The European Commission can build compute grids and federated data layers, but unless industry — from SMEs to multinational tech firms — leap to adopt and augment them, RAISE will be an academic exercise. Credit should go to the Commission for pushing this direction; but it must also create commercial levers: public procurement that uses RAISE, subsidies contingent on integration, and regulatory carrots for firms that plug in.
If the private sector views RAISE as just another bureaucratic burden or slow-moving instrument, it will steer clear — opting instead for cloud providers or proprietary infrastructures abroad. The Commission must guard against that.
One danger is that the announcement alone may be mistaken for a milestone – a mistake often made by the European Commission’s press office. It is not.
RAISE is in essence a commitment, a promise of coordination, not a breakthrough in AI itself. The real work starts now: building federated infrastructures, procuring compute, forging partnerships, attracting talent, and ensuring that Europe’s digital champions see this as a backbone, not a competitor.
Europe should not delude itself: in the new AI arms race, leadership does not reside in fragments of alignment — it resides in force multipliers, in network effects, in global adoption, in managing the “talent flywheel.” RAISE can be a part of that, but only if it is relentlessly executed.
The strategic necessity
For Europe, this is more than technology policy: it is a matter of sovereignty, economic resilience, and geopolitical posture. The next generation of AI applications — in health, transport, energy, defence — will not be neutral; they will shape who controls the invisible governance layer of global life.
By launching RAISE, Brussels tacitly acknowledges that Europe cannot remain on the margins. It must engage, build, coordinate. But ambition without clarity and resources is folly.
RAISE is a promising pivot. But its success will hinge not on the glow of its unveiling, but on the blood and iron of implementation. The European Commission must resist complacency, insist on industry uptake, guarantee cohesion among member states, and above all, invest politically rather than just financially.
If Europe genuinely wishes to move from follower to contributor in the AI era, RAISE must not become a veneer of ambition — it must become the nerve centre of a continental AI strategy. Only then will it achieve something that echoes beyond Brussels and into global influence.
Main Image: DEVILLERS Arnaud Copyright: © European Union 2014 EP Usage terms: Identification of origin mandatory
