Summary

• A Commission study prepared for the EU’s post-2027 research programme finds Europe losing ground internationally in research excellence and patenting across most digital fields.
• The EU still has more than 60,000 companies developing critical digital technologies, including 20,720 advanced digital technology startups in 2025.
• Researchers call for stronger links between research, industry, infrastructure, intellectual-property protection, demonstration projects, and commercial deployment.

Europe’s difficulty turning scientific and technical research into globally competitive digital businesses is emerging as a central problem for the EU’s next research programme, with a Commission study arguing that the bloc is losing ground in both research excellence and patenting across most critical digital fields.

The study, published by the European Commission, is intended to shape the digital component of the EU research and innovation programme under the next Multiannual Financial Framework covering 2028 to 2034. Its diagnosis is more complicated than the familiar claim that European laboratories produce strong science which industry simply fails to notice.

Instead, the researchers conclude that Europe’s weakness is also visible at the technological frontier itself. Although the EU retains scientific strengths and a substantial regulatory and industrial base, the study says it is falling behind international competitors in research excellence and patents across most digital technology areas, while remaining inconsistent at converting innovation into market success.

There is still a sizeable technology base to build from. More than 60,000 companies in the EU are developing critical digital technologies, according to the study, while the number of advanced digital technology startups reached 20,720 in 2025. Venture-capital investment across the relevant technology areas has also increased substantially since 2015.

The problem stretches beyond startup finance

Those numbers make the commercialisation problem harder to reduce to a shortage of young companies or venture money. The Commission’s study says the underlying weakness lies in the EU’s limited ability to operate consistently at the technology frontier and then carry successful research through prototyping, industrial adoption, scale-up, and deployment.

Horizon Europe has already committed around €13 billion to digital technologies through its 2021 to 2025 work programmes, particularly in microelectronics, artificial intelligence, data, and connectivity. The study nevertheless calls for deeper industrial links, better access to research and testing infrastructure, and more support around Technology Readiness Level 6, where technology moves beyond laboratory validation towards demonstration in a relevant operating environment.

That middle stage can be particularly awkward for digital technologies that are expensive to prove outside a research setting. Semiconductor work needs fabrication and packaging capacity, advanced computing needs costly infrastructure, robotics has to operate reliably around real industrial processes, and communications technology needs test networks and deployment partners before it can become a commercial product.

The policy challenge therefore extends from grants into the machinery that surrounds them. A research project can produce technically credible intellectual property without creating a European supplier capable of manufacturing, integrating, selling, supporting, and repeatedly improving the resulting technology.

AI cuts across almost every technology field

Artificial intelligence appears in the study less as a standalone sector than as an enabling layer running through several technical domains. The researchers identify links between AI and robotics, cybersecurity, advanced computing, connectivity, and edge applications, reflecting the way model development is becoming embedded in physical and digital systems rather than remaining confined to general-purpose software.

Other areas identified as promising include semiconductor-photonics integration, cybersecurity for data sharing and digital identity, advanced connectivity, and combinations of extended-reality technology with IoT and cloud-edge systems. Energy-efficient digital technology is treated as a cross-cutting priority because computing costs increasingly depend on electricity use as well as processor performance.

That last point connects research policy to the physical constraints around Europe’s broader technology strategy. Expansion of AI compute, semiconductor capacity, cloud infrastructure, and advanced networks requires electricity, grid connections, cooling, specialised facilities, and equipment supply chains, which means research funding cannot be separated neatly from infrastructure and industrial policy.

The same convergence is visible in the EU’s semiconductor agenda. Recent discussions have linked chip sovereignty with wider questions about European scale and competition, illustrating how technology policy increasingly crosses the boundaries between research grants, manufacturing capacity, market structure, and economic security.

Protecting research without closing it off

The study recommends better protection of European intellectual property and research results while retaining what it calls strategic openness towards international collaboration. That balance is difficult because advanced digital research is international by construction: universities, suppliers, standards bodies, open-source projects, and multinational companies routinely operate across borders.

Closing research ecosystems would therefore carry its own cost, particularly in fields where no single European country has the entire technical stack. At the same time, funding research whose commercial value is ultimately captured elsewhere leaves public investment disconnected from the industrial capacity that European policymakers increasingly want it to support.

The report consequently calls for more agile funding, stronger industrial anchoring, better connections between research and scale-up, expanded access to infrastructure, and continued support for energy-efficient and human-centric technologies. It also highlights open source as an area where future policy could be used more strategically.

None of those measures guarantees that a European research project becomes a European technology company of global scale. Patents can fail commercially, startups can be acquired, manufacturing can migrate, and deployment markets can favour larger incumbents even where the underlying research originated in Europe.

The study’s value lies in making the gap harder to describe as a single missing ingredient. Europe has researchers, startups, venture investment, public funding, and significant industrial demand, yet those assets do not automatically form a competitive technology system. The next EU research programme is consequently being shaped around the connections between them — laboratories to demonstrators, demonstrators to industry, intellectual property to products, and products to markets large enough to sustain them.