The Netherlands needn't get caught up in the international race for ever-larger and more expensive AI language models. That's the core message of a new deep-dive study by Invest-NL and ROM Nederland. It's not the scale of the models, but the quality of the data, energy-efficient hardware, and reliable infrastructure that determine where the next generation of AI is developed and applied. This is precisely where the Netherlands has an advantage, if we dare to choose our own AI path now.
The study, authored by AI expert Stefan Leijnen, demonstrates how global AI development is shifting toward systems that learn through sensors, robotics, and interaction with the physical world. These models must be energy-efficient and respect privacy. These are precisely the qualities that align with Dutch high-tech strengths: high-quality industrial data, strong knowledge institutions, and a robust digital and physical infrastructure.
According to Leijnen, this presents a unique opportunity for the Netherlands and Europe:
The next generation of AI isn't about even larger language models, but about models that learn through sensors, robotics, and interaction with the real world. Energy efficiency and privacy-friendliness will become key, and that's where we can make a difference.
Dutch AI opportunities in sectors where data and physical processes converge
The report identifies opportunities for the Netherlands:
- Sectors with a physical component, unique data and high adoption potential, such as science, high-tech manufacturing, healthcare, sustainable energy and agricultural technology.
- New data flows and data markets, where the Netherlands can leverage its leading position in the field of responsible AI.
- Energy-efficient AI hardware such as photonics and neuromorphic computing, areas where European and certainly Dutch companies and knowledge institutions are already recognized worldwide.
Investing in scale and less fragmentation
To capitalize on these opportunities, a robust AI industry policy is needed. The researchers emphasize that market forces alone are insufficient to achieve the necessary scale. The Netherlands must facilitate greater investment in AI startups, scale-ups, and hardware manufacturers. This means:
- Capital budgets of €20–30 million for software and applications,
- More than €100 million for hardware and foundation models.
- Stronger European cooperation to prevent fragmentation of investments, data, and infrastructure.
In addition, the government can accelerate the adoption of AI by acting as a launch customer, especially in sectors with a major social impact such as healthcare, energy, security, and scientific research.
Towards an AI Gigafactory for Europe
A striking recommendation from the study is the exploration of an AI Gigafactory: a national facility for energy-efficient edge AI, local computing power, and privacy-friendly technology. Such a strategic infrastructure could form a foundation for autonomous European AI capacity, a key element for economic growth, but also for security and the protection of European values..
Strategic course for the coming decades
Invest-NL and ROM Nederland emphasize that now is the time to choose. If the Netherlands focuses on the areas in which it is already strong, it can play a decisive international role. The report shows that a focused AI strategy is not only economically sound but essential for our future autonomy.
