Where does AI live? 10 AI companies and their workplace

AI is transforming software, finance, healthcare, logistics, industry, marketing, science and virtually every other sector. The Netherlands also has a fast-growing ecosystem of AI companies, scale-ups, infrastructure players and deeptech ventures. Substantial investment is flowing in, talent is moving toward the sector, and data centers, chips and cloud platforms are becoming increasingly decisive.

We constantly talk about what these companies build. Almost never about the buildings in which AI is built. Behind every AI organization are people, and those people need a workplace that runs on connectivity, cabling, hardware and digital infrastructure.

That is why we ask a question in this article that is rarely raised: where do the people building this AI revolution actually work, and can their workplace keep up with their digital ambitions? Put differently: if a company is AI-first, can its office still be digital-last?

Ten leading AI companies from the Netherlands

An objective top ten based on a single metric does not exist. Size, funding, valuation, revenue, growth and position in the AI chain do not all point to the same companies. That is why we deliberately speak of leading AI companies, selected for their presence in the Netherlands and relevance across different links in the chain: from infrastructure and chips to models, data, software, agents and applications.

Names and figures in the AI sector change quickly; always check the current status through sources such as Dealroom, Techleap and TNO. The selection below is intended as a snapshot and as a hook for a broader story.

1. ASML (Veldhoven). Manufacturer of lithography machines that make advanced AI chips possible worldwide. Position in the chain: the foundation of AI infrastructure.
2. NXP Semiconductors (Eindhoven). Semiconductor company focused on edge processing and connectivity. Position: chips and processing close to the application.
3. Bright Data / data players around Amsterdam. Companies that collect and prepare data for AI applications. Position: the data layer underneath every model.
4. DataSnipper (Amsterdam). AI software for auditing and finance, with strong international growth. Position: application within a specific sector.
5. Cradle (Amsterdam). AI for designing proteins, operating at the intersection of AI and life sciences. Position: models for a deeptech domain.
6. Zeta Alpha (Amsterdam). Neural search and knowledge discovery for research teams. Position: software and discovery.
7. Deeploy (Utrecht). Platform for managing and governing machine learning models. Position: MLOps and governance.
8. Orikami (Nijmegen). AI in healthcare, with applications for diagnostics and monitoring. Position: application in healthcare.
9. Slimmer AI (Amsterdam). AI venture builder that develops and scales applications. Position: building and launching applications.
10. Nebius / infrastructure providers with a Dutch presence. Providers of GPU capacity and AI cloud. Position: the compute layer on which models run.

The list mainly shows how broad and layered the Dutch AI economy is. From machines that make chips to software that applies AI: every link is anchored in the real work of people at a physical location.

And then the question IT-Label asks

We now know what these companies develop, how much capital they raise, what technology they build and how fast they grow. But what do we know about their workplace?

What digital capacity does an AI company actually need? How important is redundant connectivity, and what role does cybersecurity play? How many connected devices sit in such an office, and what happens when dozens of employees work simultaneously with heavy cloud and AI applications? Rapid growth also places its own demands on what lies under the floor and in the utility room.

The uncomfortable question behind all this: is an office building from 2010 automatically suitable for an AI company in 2026? The answer is not obviously yes. Anyone who explores the question of what a building can handle digitally will notice that surface area and rental price say nothing about it.

We still measure offices in square meters, while the organizations renting them increasingly think in data flows.

AI lives in the cloud. The employee does not.

Much AI computing power takes place in data centers and cloud environments, far removed from the office. But employees still sit in physical buildings. There they use laptops, videoconferencing, cloud software, AI agents, datasets, development environments, APIs, real-time collaboration, security tools, connected devices and business-critical SaaS applications.

AI may live in the cloud, but the connection to that cloud still begins in the building. The cloud still needs a cable. An organization is therefore only as strong as the weakest link in its digital chain: workplace, building infrastructure, connectivity, cloud and data center, and only then AI.

If one link in that chain drops out, for example because the only internet connection fails, even the smartest AI model comes to a standstill. That is why network redundancy is not a luxury for companies whose work leans entirely on connectivity, but a basic requirement.

Not every AI company has the same needs

It would be too easy to conclude that every AI company automatically needs extreme amounts of local bandwidth. A company that trains AI models has a different digital profile than a business that offers AI software through external cloud platforms. A chip developer sets different requirements than an agent company, and an organization with hundreds of engineers differs from a scale-up with thirty employees.

That is why we introduce the concept of a digital usage profile. Just as real estate professionals ask about the number of employees, square meters and parking spaces, IT-Label believes we should more often ask: how digitally does this organization actually operate?

That profile determines whether a building fits. Light office use benefits from a solid basic foundation; a data-intensive organization will look further into capacity, redundancy and management. Both are legitimate, but they call for a different way of looking at data consumption than we are used to.

From square meters to data consumption

Traditionally, we look for accommodation based on employees, workstations and square meters. In the digital economy, a second dimension is added: employees, digital processes, data consumption and infrastructure needs.

Take two organizations that both need around 1,500 square meters. A traditional administrative office with a hundred employees and an AI scale-up with a hundred developers, engineers and data scientists. On paper, the same space requirement. But does that also mean they need the same building?

Square meters tell you how much space an organization needs. They do not tell you how much digital infrastructure that organization needs. That second question usually remains unanswered today, even though for digitally intensive users it can weigh more heavily than the rental price per square meter.

What is an AI-proof workplace?

The term AI-proof office or AI-ready workplace is not an official, universal certification. Use it as a conceptual question: can this workplace keep pace with an organization that runs entirely on data and cloud?

A future-proof digital workplace requires attention to a number of interconnected themes:

What an AI-ready workplace looks at

• Connectivity: fiber, capacity, reliability and redundancy.
• Internal infrastructure: data cabling, WiFi, network equipment and technical spaces.
• Cybersecurity: physical access, network segmentation, clear responsibilities and management.
• Continuity: backup connections, power supply and incident procedures.
• Smart building: sensors, IoT, building data, access control and integrations.
• Scalability: can the digital infrastructure grow along as the organization grows?

The common thread is simple. An AI-proof organization deserves an AI-ready workplace. Topics such as fiber in buildings and reliable WiFi coverage are not technical details, but load-bearing elements of daily productivity.

What does an AI company actually receive digitally?

When a fast-growing AI company rents an office, it receives extensive information about floor area, rental price, service costs, energy label, parking spaces, accessibility and delivery level. The question is whether it also gets an answer to the things that affect its business model.

What fiber connection is present, and which providers are available? Is there redundancy? What data cabling and WiFi infrastructure are in place? What is mobile coverage like? What technical space is available, which smart building systems are connected, who manages what, and what must the tenant arrange themselves?

We know exactly how many square meters an AI company rents. But do we also know what those square meters can handle digitally? That information is still almost always missing today, even though it can be decisive in a digital due diligence.

The role of IT-Label

IT-Label is an independent knowledge collective and a classification methodology that makes the digital delivery level of commercial real estate transparent. The classification runs from IT1+ PREMIUM and IT1 HIGH PERFORMANCE through IT2 PLUG & PLAY and IT3 READY to IT4 CORE and IT5 SHELL.

IT-Label does not state how much AI an organization can run, nor does it guarantee that a building is suitable for every AI company. It makes the digital foundation of the property transparent, so that users and landlords can better determine whether it matches the actual need.

The user knows what they need. IT-Label makes visible what the building delivers digitally. This makes the conversation between tenant and landlord more concrete, without either party having to persuade the other based on impression alone.

The next location question of an AI company

Today, a company asks a broker: how many square meters, what is the rental price, how many parking spaces, what is the energy label? Tomorrow, an additional question will increasingly be added: what can this building handle digitally?

Especially for companies whose business model relies entirely on data, cloud and AI, digital infrastructure is becoming more relevant within the accommodation decision. But this is not only about AI companies. They are the frontrunners of a much larger development.

Law firms, accountants, banks, logistics companies, healthcare organizations, real estate agents, retailers and government bodies are also becoming increasingly digital. Today, AI companies set these requirements. Tomorrow, they may simply be the requirements of every office. For property owners, that is a reason to look at the digital foundation now.

AI does not only change software, AI changes real estate

AI changes how organizations work. How organizations work determines what they need from their workplace. And what they need from their workplace ultimately determines which buildings remain attractive.

The reasoning builds step by step. More AI leads to more digital processes. More digital processes lead to a different way of working. A different way of working leads to a different digital need, and thus to different requirements for the workplace and ultimately for the real estate itself.

The AI revolution may take place digitally, but its consequences ultimately also become visible in square meters. Anyone who wants to build or manage for the future cannot ignore that chain.

Real estate value in the digital economy

If technology companies and other digitally intensive users start setting higher requirements for digital infrastructure, a bigger question arises: can digital quality become part of the appeal and value of real estate?

Some caution is warranted here. We do not claim that a higher IT-Label automatically leads to a higher rental price or property value. There are, however, reasonable questions worth exploring. Can better digital infrastructure reduce vacancy risk? Can transparency speed up leasing? Can a digitally better prepared building be more attractive to tech and AI companies? And could digital quality in the future become part of due diligence and valuations?

In an economy increasingly driven by data, the infrastructure that makes data possible naturally becomes more important. That insight affects both the rental market and the way we will assess the value of real estate in the future.

Frequently asked questions

What are the most important AI companies in the Netherlands?

There is no single objective top ten, because size, funding and position in the chain point to different companies. The sector spans everything from chip machines and semiconductors to models, data, software and applications. Always check the current status through sources such as Dealroom, Techleap and TNO.

Does an AI company need a special office?

Not every AI company has the same requirements. A company that trains models has a different digital usage profile than an organization that offers AI software through the cloud. The need depends on the work, not just the number of employees.

What is an AI-proof or AI-ready workplace?

This is not an official certification, but a conceptual question about connectivity, internal infrastructure, cybersecurity, continuity, smart building systems and scalability. In short: can the workplace grow along with a digitally intensive organization?

What exactly does the IT-Label make transparent?

The IT-Label makes the digital delivery level of a building transparent, from IT1+ PREMIUM to IT5 SHELL. It does not judge the building as good or bad, but shows what the digital foundation is, so that users and landlords can weigh the actual need.

Does this only apply to AI companies?

No. AI companies are the frontrunners. Law firms, banks, healthcare organizations and government bodies are also becoming increasingly digital, which means the requirements AI companies set today may eventually become the requirements of nearly every office.

So where does AI actually live?

The Netherlands is building impressive AI technology. We invest in chips, models, cloud, agents, data and talent. Perhaps we should also look at the places where that talent works every day.

We know which companies are building the Dutch AI revolution. The next question is whether our buildings are ready to house that revolution. AI changes our companies, AI changes our workplace, and ultimately AI changes our real estate.

A concrete first step is to gain insight into the digital foundation of your property. Discover how you can ensure the digital foundation of your real estate is right and which classification fits your building. What can your building handle digitally?