AI adoption is stalling in many Dutch organisations due to a lack of knowledge, experience and ownership. According to research of InSpark, 59% of the organisations surveyed identify a lack of knowledge as their biggest obstacle. The technology is available, but applying it in everyday work requires direction, guidance and a shared learning process.
Summary

Dutch organisations are experimenting extensively with AI, yet widespread adoption is still lagging behind. Only 8% use AI across the organisation and have embedded it strategically. Most organisations are still exploring, testing or waiting to see what happens.
This points to a shift in the challenge. Access to technology is becoming less of an issue. The real challenge is developing the ability to use AI responsibly, practically and in ways that create value.
Teams need knowledge, leaders need direction, and organisations need clear agreements around quality, security, roles and ownership. AI only becomes valuable when people learn how to use it within the context of their own work.
That makes AI adoption primarily an organisational and learning challenge.

Key insights

• 59% identify a lack of knowledge as the biggest obstacle. AI adoption therefore requires more than tools and licences.
• 8% use AI strategically across the organisation. Most organisations are still exploring, running pilots or using AI in isolated areas.
• 25% use AI structurally across multiple processes. Value is already being created, although often without organisation-wide adoption.
• 31% are primarily exploring the possibilities. Exploration can be valuable, as long as it leads to clear choices and next steps.
• AI affects roles, processes and decision-making. Its impact remains limited when teams only learn to use tools without developing new ways of working.

The pilot phase is lasting too long
Many organisations have already started using AI. Employees use ChatGPT, Microsoft Copilot and other applications. Sometimes this is centrally organised, while in other cases enthusiastic colleagues are independently discovering what is possible. That energy is valuable. It shows that AI is no longer an abstract topic for the future. The technology has reached the workplace. New applications are emerging every day across sales, HR, finance, operations, product development and communications.

At the same time, a familiar pattern is emerging. There is interest, there are experiments, and there are examples of time savings or better preparation. Yet the step towards consistent, organisation-wide use remains difficult. One department moves quickly while another is still looking for guidance. Some teams use AI every day, while others are waiting for policies, training or approval.

According to InSpark, only 8% of the Dutch organisations surveyed use AI strategically across the entire organisation. A quarter use AI structurally within multiple processes. The remainder are primarily exploring, running small-scale pilots or are not yet using AI at all. This reflects what we see in our own conversations with organisations. AI itself is rarely still up for debate. Applying it effectively in practice is where the challenge lies.

Knowledge is the first bottleneck
The most striking figure from InSpark’s research is 59%. That is the percentage of organisations that identify a lack of knowledge and experience as the biggest barrier to AI adoption. That figure deserves attention. It suggests that the main constraint on AI adoption lies in people, application and organisational design. The tools are often already available. The interest is there too. What is missing is enough understanding to use AI responsibly and purposefully.

And knowledge goes far beyond knowing how to write a prompt. Employees need to understand what context is required to generate useful output, how to assess results and where risks can arise. Teams need to learn when AI can help and when human judgement should remain decisive. Leaders need to determine where AI can contribute to strategy, service delivery, efficiency or quality.

Without that knowledge, fragmentation emerges. One employee develops an effective way of working, while another enters sensitive information into an AI tool without fully understanding the risks. One team saves time on preparation, while another repeatedly makes the same mistakes because there is no shared learning structure. AI adoption therefore requires organised learning. That means more than an introductory session. Organisations need a rhythm in which people practise, share experiences, develop quality standards and connect AI applications to their own work.

Ownership falls between departments
AI affects several areas of an organisation at the same time. As a result, ownership is often unclear. IT focuses on security, data, architecture and tooling. HR and L&D focus on skills, training and adoption. Legal and compliance focus on risk and regulation. Business teams focus on productivity, customer value and process improvement. Leadership focuses on strategic impact and investment decisions.

All of these perspectives are necessary. And that is precisely why AI can end up falling between departments. When AI is primarily treated as software, the focus shifts towards licences, management and policies. When it is mainly treated as a learning and development challenge, the link with strategy and operations can remain too weak. When AI sits solely with innovation teams, promising experiments can emerge without a clear path towards broader adoption.

Strong AI adoption requires clear ownership at multiple levels. Leadership needs to provide direction, teams need practical support, and sufficient technical and legal safeguards must be in place. It should also be clear who makes decisions about priorities, budgets, risks and scaling. Without those agreements, AI remains driven by enthusiasm. With clear ownership, it can develop into a mature organisational capability.

Funding often comes too late Funding for AI development is still unclear in many organisations. This slows down the move from experimentation to adoption. AI training rarely fits neatly into a single budget category. It is partly technology, partly learning and development, partly organisational change and partly strategic innovation. As a result, discussions often arise about which budget should pay for it.

That may sound like a practical issue, but it has strategic consequences. When funding is only arranged after a pilot, delays occur at exactly the moment when momentum and evidence are available. Teams that are ready to move forward suddenly have to return to the budgeting process. It helps to decide in advance what type of investment AI adoption represents. Is it about digital skills? Process improvement? Productivity growth? Risk management? Strategic innovation?

In practice, it will often be a combination of these. That is exactly why an explicit choice is necessary. Not every AI initiative requires a large budget, but every serious adoption programme requires capacity, guidance and decision-making.

Area	Typical owner	Risk when ownership is unclear
Tooling en security	IT	Usage grows without clear guardrails
Skills and training	HR and L&D	Employees experiment without sufficient guidance
Process improvement	Business teams	AI remains disconnected from tangible value
Governance and risk	Legal, compliance and leadership	Decisions are delayed or made too late
Scaling	Leadership and transformation teams	Pilots generate insights but fail to scale

This division of responsibilities shows why AI adoption does not naturally fit within existing organisational structures. Organisations that make greater progress are explicit about who plays which role.

AI changes work, roles and collaboration
AI adoption affects more than individual productivity. As AI is applied more widely, tasks, roles and ways of working begin to change. An employee who uses AI for analysis, preparation or documentation works differently from before. A team that uses AI during refinement, customer research or reporting needs new agreements around quality and responsibility. A leader using AI to support decision-making or scenario planning needs to pay closer attention to interpretation, bias and context.

At that point, the impact shifts from tool usage towards organisational design. Work is redistributed. Tasks disappear, change or move between roles. Teams gain new capabilities, while also developing new dependencies. Leaders need to guide people through those changes. In that sense, AI adoption resembles previous major shifts around Agile, DevOps and product-led ways of working. The methodology or technology is highly visible, but the real work happens in behaviour, collaboration and application.

Organisations that view AI primarily as an efficiency tool are only seeing part of the opportunity. The greater value lies in learning better, thinking faster, collaborating more effectively and making sharper decisions.

Training only works when practice surrounds it
Training matters, but training alone is too limited to enable widespread AI adoption. An introduction can help people get started. A workshop can provide examples. A programme can create structure. Then the real learning begins.

That is where the important questions emerge. Which prompts work within our processes? Which outputs are suitable for our customers? What information can we safely enter into a tool? How do we assess quality? How do we share successful applications across teams? How do we prevent everyone from independently reinventing the wheel?
Answering these questions requires a learning environment around the work itself. People need opportunities to experiment, exchange experiences and refine how they use AI. Leaders need to create room for learning without requiring every experiment to immediately justify itself through a business case. Teams need guardrails that provide direction without restricting everything.

At Connected Movement, we see training, guidance and community reinforcing each other in this process. Training provides a foundation and a shared language. Guidance helps people apply what they have learned within their own context. Community keeps the learning going by allowing professionals to compare their experiences with others facing similar challenges. AI adoption then becomes a way to strengthen the organisation’s overall capacity to learn.

Moving from intention to action
Many organisations want to get started with AI. The challenge lies in turning that ambition into manageable steps.

An organisation does not need to wait until it has developed a complete AI strategy for every department. A better first step is often smaller and more concrete: choose a recognisable work process, bring together a mixed group from business, IT, HR and risk, establish clear guardrails and learn in short cycles what works. The first applications should be safe enough to experiment with and relevant enough to demonstrate value. Examples include internal knowledge sharing, preparing customer conversations, analysing feedback, documentation, reporting or improving recurring tasks.

Those initial applications begin to provide evidence. Where are we saving time? Where is quality improving? Where do risks emerge? Which skills are missing? Which agreements need to be made across the organisation? This allows AI adoption to grow from practical experience rather than abstract ambition.

Conclusion
InSpark’s research shows that AI adoption in the Netherlands is being held back less by technology and more by knowledge, experience and organisational capability. That is an important insight because it means organisations can actively address the problem. Making tools available is relatively straightforward. Helping people use AI professionally, safely and in ways that create value requires more attention. It takes clear ownership, targeted development, practical guidance and leadership.

The organisations that make progress treat AI adoption as a learning process. They start small, establish agreements around quality and risk, connect applications to real organisational goals and build confidence step by step. AI then stops being a separate experiment alongside everyday work. It becomes part of how people collaborate more effectively, learn faster and create value more consciously.

Next step

Make AI adoption concrete
Want to understand where AI adoption is getting stuck in your organisation? Start with one work process, one team and one clear application. Explore what knowledge, agreements and guidance are needed to create value safely.

Book an exploratory conversation Author

Wesley van de Pol
Sales Director at Connected Movement. Wesley speaks with organisations every day about strengthening their ability to change across Agile, digitalisation and AI. In his role, he connects client challenges with the right expertise, guidance and practical next steps.