Summary
- EU R&D spending reached 2.24% of GDP in 2024, remaining below the bloc’s 3% target and several large competitors.
- The Commission identifies relative European weaknesses in AI, advanced semiconductors, cloud computing, and commercial scaling.
- Finance, fragmented markets, technology diffusion, and skills remain obstacles between research capability and productivity growth.
Europe retains one of the world’s strongest scientific bases but continues to lose momentum between research, commercial deployment, and industrial scale, according to a European Commission assessment that places the continent’s technology problem less in invention than in what happens afterwards.
The 2026 Science, Research and Innovation Performance report from the European Commission examines that gap through what it calls strength, structure, and scale. European universities and research organisations continue to produce substantial scientific output, yet investment, market fragmentation, and uneven commercialisation make it harder for technologies to develop into large businesses and spread through the economy.
Research and development spending across the EU rose from 1.81% of GDP in 2000 to 2.24% in 2024, but remains below the bloc’s longstanding 3% target. The Commission compares that with 2023 R&D intensity of 3.45% in the US, 3.44% in Japan, and 4.96% in South Korea.
The composition of investment also differs. European public funding remains strong in universities and basic research, while private R&D is relatively concentrated in established manufacturing industries rather than the faster growing technology and knowledge intensive sectors absorbing much of the current global investment in AI and digital infrastructure.
Scientific strength is not translating evenly into industry
The EU ranks second globally for scientific output, while 57% of European publications involve international or cross-border collaboration. That network provides a sizeable base of expertise for industries that depend on specialist science and long development cycles, but Europe’s position weakens as research moves closer to technology markets.
Its share of international patent applications under the Patent Cooperation Treaty fell from roughly 30% in 2000 to around 16% in 2022, while China’s share rose from about 3% to almost 33%. Patents are an imperfect measure of industrial strength, although the change adds to evidence that economic weight in several technology sectors has moved eastwards and towards the US.
The Commission still identifies European strengths in advanced materials, manufacturing, and several technologies associated with the green transition. Digital technology presents a less favourable picture, particularly in artificial intelligence, advanced semiconductors, and cloud computing, where European businesses often rely on infrastructure and suppliers headquartered elsewhere.
EU innovation output has increased by almost 10% over the past decade, but knowledge transfer and commercialisation differ markedly between countries, institutions, and sectors. Collaboration between companies and research organisations also remains uneven, leaving promising technology vulnerable to losing momentum once a project has to find customers, investment, manufacturing capacity, and experienced management.
Productivity reflects part of that weakness. The Commission estimates that EU productivity growth over the past 25 years has run at roughly half the US rate, with fragmented markets, regulatory complexity, weaker business dynamism, and slow diffusion of technology contributing to the difference.
Scale requires more than research grants
European venture capital has expanded substantially, but later stage finance remains a persistent weakness. Institutional investors in the EU collectively control around $12 trillion of assets yet commit comparatively little to venture investment, while roughly half of the growth stage funding raised by European deep technology companies continues to come from outside Europe.
Public programmes can attract private money into earlier stages. The Commission says startups supported by Horizon Europe and earlier research programmes have subsequently raised €70 billion in venture capital since 2010, while the European Innovation Council has mobilised more than €4 billion in additional private investment.
Those figures do not remove the financing problem once a company needs to manufacture hardware, build infrastructure, hire internationally, or enter several markets at once. Businesses that depend on non-European capital at that stage may scale successfully, but ownership, intellectual property, senior management, or future investment can migrate with the funding.
Artificial intelligence brings several of those constraints together. Europe’s share of global AI patent applications fell from around 20% in 2000 to 12% in 2022, even though major European cities remain competitive in AI research. The region is now spending heavily on AI factories, semiconductor projects, computing capacity, and data infrastructure in an attempt to strengthen the machinery between scientific work and commercial deployment.
Computing capacity alone will not close the gap. Companies still need customers willing to adopt new systems, technical employees capable of implementing them, markets large enough to reward expansion, and finance able to carry a business through expensive growth stages. A new AI factory can supply compute, but it cannot by itself create demand or solve fragmented procurement.
The report therefore lands as European technology policy becomes more industrial in character. Research funding remains central, although governments are increasingly concerned with energy, data centres, manufacturing, procurement, capital markets, and the ability to retain companies after their technology proves viable.
Europe’s scientific institutions continue to supply much of the knowledge required for new industries, while the weakness lies in the economic system surrounding that knowledge. Closing the gap will depend on whether financing, infrastructure, markets, and adoption improve quickly enough for research developed in Europe to be deployed there before competitors turn similar ideas into larger businesses.










