
Deep tech
—27 november 2025
November 27, 2025
4 minutes

The Netherlands can and does not need to compete with the United States and China in the race for ever larger and more expensive AI language models, nor should it want to copy that approach. This is demonstrated by a new deep-dive study by Invest-NL in collaboration with ROM Nederland. The research shows that the Netherlands is especially strong in those parts of the AI chain where qualitative data, energy-efficient technology, reliable infrastructure, and specialist knowledge are decisive. To capitalize on this position, investments and a robust AI industry policy are necessary.
The study, authored by AI expert Stefan Leijnen, analyses how global AI development is shifting towards new energy-efficient hardware and new data streams and applications that are strongly rooted in physical processes and industrial data. According to the report, the Netherlands can become internationally distinctive in these areas if it makes strategic choices now.
"The next generation of AI isn't about even larger language models, but about models that learn via sensors, robotics, and interaction with the real world. These models will need to be energy-efficient and privacy-friendly," says author Stefan Leijnen. "There are opportunities for the Netherlands and Europe there."
The deepdive demonstrates that the Netherlands is well positioned where physical processes, high-quality data, and high adoption potential come together: agricultural technology, logistics, high-tech manufacturing, energy, and healthcare. In these sectors, AI applications that are difficult to copy are emerging, which run on protected data and are integrated into the physical world.
The report identifies opportunities for the Netherlands:
The study advocates for an AI industry policy that removes barriers and invests strategically where market mechanisms fall short. This policy should focus on increasing access to capital for AI startups and scale-ups, with tickets of €20–30 million and more than €100 million for hardware and foundation models, combined with closer European cooperation to reduce fragmentation in investments, data, and infrastructure.
Additionally, the government should use public demand to accelerate adoption, especially in domains such as healthcare, energy, research, and security, and build public institutions that bring vital data and digital infrastructure under democratic governance.
Finally, the study calls for research into a future-proof AI-Gigafactory for edge-AI, energy-efficient hardware, and privacy-conscious local computing power, as a strategic foundation for autonomous AI capacity in the Netherlands and Europe.
Invest-NL and ROM Netherlands emphasize that strategic investments are crucial at this moment to strengthen this position. “If the Netherlands chooses its own AI course, we can succeed particularly in the areas where we are strong. Our investment policy should focus on larger tickets for AI startups and hardware manufacturers, investing where the market falls short, and connecting parties around a strategic long-term agenda. This deepdive shows exactly where those opportunities lie.”
The report illustrates that a Dutch AI strategy is not only economically wise but essential for future security, autonomy, and protecting European values. The AI deepdive follows earlier publications from Invest-NL on quantum, lab-on-a-chip, and semicon – all key technologies within the National Technology Strategy. The deepdive series aims to show how and where targeted investments create strategic advantage for the Netherlands.
Liz Duijves
investment manager
