The world's most generous incubator

A quiet shift is underway across Europe, driven by three relentless forces: demographic decline, a fractured geopolitical order, and rapid technological transformation. Over the next decade, these forces will require Europe to rethink its economic model, potentially bringing about the most significant changes since the creation of the Single Market. Whether Europe survives as an industrial superpower depends critically on how fast and how precisely we align our greatest asset - our human capital - with the capability to build and scale advanced technologies.

For the moment, the numbers point to a persistent European challenge. Every year, the European Union and the United States produce comparable numbers of STEM doctoral graduates: the US produces approximately 44,700 STEM PhDs annually [ref 1], compared to roughly 40,000 to 45,000 across the EU [ref 2], while Europe generates nearly a quarter of the world’s top-cited scientific publications (with just 6% of the world’s population) [ref 3]. Yet, the moment those brilliant minds attempt to turn their discoveries into global industries, the trajectories diverge dramatically.

Even though European early-stage startups survive at comparable rates to American ones, a European company is reportedly 40% less likely to secure venture backing after five years. At the same time, Europe captures just 14% of global late-stage growth funding ($100M+ rounds), while the US commands 50% [ref 4]. At that point, up to 20% of our highest-potential scale-ups execute what is - rather tragicomically - called a “Delaware Flip”, relocating their legal headquarters and intellectual property across the Atlantic [ref 5]. Furthermore, nearly a third of European Deep Tech exits (where startups built on hard science or engineering are sold or listed on the stock market) end up acquired by US buyers [ref 6]. These patterns raise important questions about Europe's ability to reap the benefits of its research and innovation strengths.

For years, discussions about this challenge have focused on whether Europe lacks "entrepreneurial spirit", "oomph", or even the money. To put it bluntly, Europe too often functions as a taxpayer-funded R&D laboratory and talent incubator for external capital markets. And as we debate defence imperatives or industrial disruption, one foundational truth is becoming increasingly clear for European policymaking: supercomputers, fabs, and industrial funds do not innovate. People do.

Which brings us to the obvious question: if people innovate, how are we actually preparing them?

Right now, there is a lot of talk in Brussels about the Union of Skills. Don't get me wrong, it is a vital initiative. But if we focus merely on addressing immediate skills shortages, without investing in the development of longer-term capabilities, we are missing the point entirely. It also raises a more fundamental question: how do we know which skills Europe will need? Labour market statistics and employer surveys provide valuable evidence about existing shortages and recent trends. For many sectors, these are a useful guide for policy. But can we apply the same logic to AI? In areas where technology is moving this fast, the recent past can only tell us so much. We cannot prepare people for tomorrow’s economy by looking only in the rear-view mirror.

We need to combine labour market evidence with much closer engagement with those working at the frontier of science and technology. Universities, researchers and innovators can help us understand emerging technologies and the capabilities needed to shape them, long before demand appears in official statistics.

Today, only 13.5% of EU enterprises - and a mere 6.9% of our SMEs - have integrated AI into their daily operations [ref 7]. We will not win tomorrow’s race simply by training Europeans to be slightly more efficient users of software built in California. We also need to educate the people who will invent the industries that do not even exist yet.

To do that, we must look in the mirror first.

The growing political attention to STEM education is welcome. Everybody is talking about the engineering shortage and the need for more advanced technological skills. But have we really recognised, at European and national level, what it takes to deliver on these ambitions? Educating more scientists and engineers, while maintaining excellence, requires more educators, laboratories, infrastructure and support for students. We cannot keep raising our ambitions for STEM education without matching them with sustained investment in the universities expected to deliver.

And look at AI. It is already transforming how we teach, learn and assess knowledge, raising fundamental questions about the future of education. Universities of science and technology are at the forefront of responding to these changes, experimenting with new approaches and sharing experiences and evidence across Europe, including through CESAER. Yet too much of the responsibility still falls on individual professors, who need time, support and resources to adapt their teaching and assessment. At the same time, unequal access to high-quality AI tools and infrastructure risks creating new divides between students, universities and countries. Europe needs to treat secure, equitable and responsibly governed AI infrastructure as a strategic investment.

Crucially, this is not just about AI - whatever we make of the hype surrounding it - it is about how we respond to transformative technologies and prepare the next generation to shape them. We should lead continuous, evidence-based experimentation, building deep technological literacy while fiercely protecting what no algorithm can replace: rigorous mathematical foundations, critical thinking, and independent human judgement. If we want universities to lead this transformation, we must give them the means to do so.

As the European Skills High-Level Board shapes its agenda and negotiations heat up around the next Multiannual Financial Framework (MFF), I believe we need three decisive shifts:

1. Establish an ERA-wide IP Charter and clearer, more consistent frameworks for university equity stakes in spin-offs, paired with a unified "28th Regime" for startups, while ensuring that academic career assessment recognises entrepreneurship and knowledge valorisation alongside scientific publications.
2. Europe has €33 trillion in household savings [ref 8], much of it held in low-yield deposits, while pension funds allocate just 0.007% to venture capital (less than a quarter of the US rate). Unlocking even a fraction of that capital for university-anchored deep tech scale-ups could significantly strengthen Europe's capacity to scale its own innovations.
3. Invest in universities of science and technology as strategic infrastructure. Through the next MFF and sustained national investment, strengthen teaching and research capacity, European academic computing infrastructure, university-industry learning environments and shared fabrication labs. Let’s turn talent-rich regions across the Union from outsourcing hubs into centres of deep tech innovation and industrial development.

The talent is already sitting in our amphitheatres. The capital is already sitting in our banks. The choices we make right now will determine whether we finally connect the two, turning our scientific and technological strengths into the innovations and industries of the future. 

By Mihnea Costoiu, Rector of the National University of Science and Technology POLITEHNICA Bucharest and CESAER Envoy to the European Skills High-Level Board

References

Ref 1: National Science Board, National Science Foundation, "International Comparisons and Global Competitiveness," in STEM Talent: Education, Training, and Workforce, Science and Engineering Indicators 2026, NSB-2026-1, Alexandria, VA, USA: NSF, Feb. 2026. [Online]. Available: https://ncses.nsf.gov/pubs/nsb20261/international-...

Ref 2: Eurostat, "Graduates by education level, programme orientation, sex and field of education (educ_uoe_grad02)," European Commission, Luxembourg, Updated Oct. 6, 2026. [Online]. Available: https://ec.europa.eu/eurostat/databrowser/product/...

Ref 3: Directorate-General for Research and Innovation (European Commission), "Science, Research and Innovation Performance of the EU (SRIP) report," European Commission, Oct. 2026. [Online]. Available: https://research-and-innovation.ec.europa.eu/strat...

Ref 4: Atomico, "The State of European Tech," Atomico, London, U.K., Nov. 2025. [Online]. Available: https://stateofeuropeantech.com/

Ref 5: G. Testa, R. Compano, A. Correia, and E. Rückert, "In search of EU unicorns - What do we know about them?," Joint Research Centre, European Commission, Luxembourg, Rep. EUR 30978 EN, JRC127712, 2022, doi: 10.2760/843368. [Online]. Available: https://publications.jrc.ec.europa.eu/repository/h...

Ref 6: C. Fratto, M. Gatti, A. Kivernyk, E. Sinnott, and W. van der Wielen, The Scale-Up Gap: Financial Market Constraints Holding Back Innovative Firms in the European Union, European Investment Bank, Luxembourg, Rep. QH-09-24-635-EN-N, Jul. 2024, doi: 10.2867/382579. [Online]. Available: https://www.eib.org/en/publications/20240130-the-s...

Ref 7: Eurostat, "Digitalisation in Europe – 2025 edition," Interactive Publications, European Commission, Luxembourg, 2025. [Online]. Available: https://ec.europa.eu/eurostat/web/interactive-publ...

Ref 8: E. Letta, Much More Than a Market: Speed, Security, Solidarity - Empowering the Single Market to Deliver a Sustainable Future and Prosperity for All EU Citizens, High-Level Report to the European Council, Apr. 2024. [Online]. Available: https://www.consilium.europa.eu/media/ny3j24sm/muc...