The government's strategic choice in AI innovation

The Dutch model of laying foundations

The Dutch approach, rooted in a philosophy of building sustainable, broadly accessible infrastructure, is bearing increasing fruit. It goes beyond merely providing facilities; it creates a culture of collaboration and applied innovation. This is visible in practice.

A telling example is longform.ai, a spin-off from the University of Amsterdam that emerged in 2024 within the ICAI Discovery Lab [7]. This startup, which is developing an AI platform for audio analysis, immediately raised seed capital from the Innovatiefonds Noord-Holland. It is a textbook example of how the infrastructure of public-private labs directly gives rise to commercial and fundable companies.

This approach is not limited to software. The Dutch "meta-tech" mentality, not making the chips, but making the machines that make the chips, is also visible in AI. Companies such as QuantWare (quantum computing) and Fizyr (robotics vision) are building complex systems that integrate AI. They benefit from close collaboration between technical universities, government and specialised investors, a dynamic reinforced by platforms such as Techleap and the Dutch AI Coalition [8].

The French model of forging champions

France has opted for a significantly different, more interventionist strategy. Since 2018, the French government has committed more than €2.5 billion to an ambitious AI plan aimed at creating European leaders [5]. This policy combines infrastructure investment with highly targeted, direct support for promising startups.

The most prominent example is Mistral AI. Although the French state is an important indirect shareholder through public funds, the recent funding round of €1.7 billion was led by private parties, with Dutch company ASML as the largest investor, contributing €1.3 billion. This approach has delivered clear results. In 2024, French AI startups raised €1.9 billion and the country was the most popular destination in Europe for foreign AI investment [5].

The innovation path for startups and scale-ups

For entrepreneurs, the government's strategic choice makes a world of difference. The Dutch approach provides fertile ground for early-stage innovation, while the French route can offer a faster path to the top for a select group.

For the early-stage startup in the Netherlands The infrastructure model is a genuine asset. An early-stage company can take advantage of the following benefits.

• Validate without major capital investment. Access to the computing power of the AI Factory or the expertise of an ICAI lab lowers the barrier to testing complex models.

• Attract talent. The close ties with universities through ICAI create a steady stream of highly educated PhD graduates and students.

• Collaborate on R&D. Participating in an ICAI lab offers the opportunity to work alongside academics and larger companies on fundamental breakthroughs.

For the scale-up in the Netherlands This is where the challenge grows. Once a company has moved beyond the fundamental research phase and needs to scale, the Dutch model offers less direct support. The strategic choice at that point is as follows.

• Seek international partners. The investment by Dutch company ASML in French Mistral AI is a perfect example of this. Scale-ups must proactively pursue European capital and strategic partnerships.

• Focus on a profitable niche. Rather than competing on pure scale, Dutch scale-ups can target highly specialised, profitable B2B markets where their technological edge makes the difference.

Sources: ICAI.ai [3], Business France [5], IO+ [2]

The road ahead: from foundation to global player

With its infrastructure model, the Netherlands has laid a solid and fertile foundation. The next step is not to replace this model, but to complete it. The challenge now is to build deliberately on this foundation, so that tomorrow's successes not only germinate here, but also grow into global players here. For startups, there are concrete opportunities to be found.

1. AI for the energy transition The Netherlands faces an enormous challenge in the energy transition. Startups such as Dexter Energy, which has raised €23 million in funding and uses AI to forecast and optimise energy flows, demonstrate that a substantial market exists here [9]. The combination of strong energy infrastructure and AI expertise is a unique asset.

2. AI in the 'old' economy The strength of the Dutch economy lies in sectors such as logistics, agriculture and the maritime industry. Startups that apply AI to optimise processes in these sectors have a direct competitive advantage. They solve real problems for a capital-rich market that is eager for efficiency gains.

3. Deep tech and system integration Building on the 'meta-tech' mentality, opportunities exist in the complex integration of AI with other advanced technologies such as quantum computing and photonics. This is not a market for quick returns, but a domain where the Netherlands, with its patient, collaboration-oriented approach, can build a sustainable advantage.

To capitalise on these opportunities, an evolution in government policy is desirable. In addition to maintaining the infrastructure, a logical next step would be to demonstrate greater strategic ambition, following the French model. This could take the form of co-investment funds specifically targeted at the scale-up phase, bridging the gap between early-stage innovation and the large international capital market. In this way, the fertile soil would be used not only for sowing, but also for harvesting the most promising crops.