In conversation with Kees Veenhof, COO and director of legal services at ARAG Netherlands, it becomes clear that the impact of AI is ultimately not determined by the technology itself, but by how leaders make choices, set priorities, and bring their organisation along through change.
From fascination to focus: where does value begin?
When generative AI arrived, a period of exploration began at ARAG too. The technology was intriguing, but it quickly became clear that experimenting without direction produces little. Real movement only started once AI was tied to a fundamental question: how can we improve our service while keeping it affordable?
That focus on customer value creates sharpness. AI didn't become a goal in itself, but a tool to speed up service, improve processes, and keep costs under control. That combination turns out to matter. Organisations that treat AI as separate from their core processes and customer promise often end up stuck with scattered initiatives that don't produce lasting impact.
"We shouldn't do something with AI — we should solve something AI can actually help us with."
Not an IT topic, but a boardroom topic
One of the most pointed insights from the conversation is that AI belongs on the executive agenda. Not as a technical experiment, but as a strategic question with direct influence on how business strategy gets executed.
That seat at the table asks something in return. It means progress isn't just discussed occasionally, but monitored structurally. That there's clarity about what initiatives are supposed to deliver. And that there's ongoing attention to how AI contributes to the goals the organisation sets for itself. At the same time, attributing impact cleanly turns out to be difficult; different initiatives affect one another, so success is rarely tied to a single factor. Even so, that conversation about value and effect is exactly what keeps focus in place.
Start small, think big
Rather than a broad rollout, ARAG chose a focused starting point with a concrete application: generating legal plans of action for clients. That tied AI directly to a core process.
What stands out is that the biggest acceleration didn't come so much from the technology itself, but from how the work was organised. By minimising wait times between steps and putting together multidisciplinary teams, a much faster development rhythm emerged. New versions could be delivered in weeks instead of months.
Organising feedback also proved decisive for success. Giving users direct influence over how the solution develops creates ownership. But that only works if feedback gets processed and made visible quickly. Once that feedback loop is missing, trust disappears, and with it the motivation to contribute.
AI as a mirror: what isn't working becomes visible
A recurring insight is that AI exposes existing weaknesses without mercy. Processes that were already inefficient don't get fixed by technology — they get amplified.
That's why ARAG's priority is explicitly on getting processes and data in order first. Only once the foundation is right can AI actually add value. The alternative is that organisations create speed in the wrong direction. That might sound efficient, but it ultimately leads to more complexity and higher costs.
"Don't use AI to speed up an inefficient process — you'll just get to the wrong place faster."
The new customer: better prepared, not always better informed
AI is changing not just the internal organisation, but the customer relationship too. Customers increasingly arrive with advice they've already generated themselves. That changes the conversation.
Where professionals used to start straight in with analysis and advice, the conversation now more often starts with correcting or nuancing AI-generated insights. Customers are prepared, but not always fully or correctly informed. That calls for extra attention, more explanation, and a different way of communicating.
The professional's role shifts as a result. It's not just expertise that counts, but also the ability to manage expectations and make complex information understandable. AI adds an extra layer to the interaction — one that isn't technical, but human.
Adoption is the real challenge
As in many organisations, technology turns out to be only part of the story here too. The real challenge is in adoption.
A familiar pattern shows up within ARAG. Some employees embrace the new possibilities and experiment actively. Others hold back and wait to see how things go. And there's always a group that drops off once early experiences disappoint.
That first experience turns out to matter most. If a solution doesn't fit the work in practice right away, or if improvements don't follow, people quickly go back to their familiar way of working. That's why the emphasis is on creating positive first interactions, supported by fast improvement cycles and involvement from subject-matter experts. Requiring people to use it works less well than making them want to. The value has to be felt in day-to-day work.
Between transparency and speed: a deliberate choice
In an environment where legal accountability is central, explainability isn't a luxury but a precondition. That makes the trade-off between performance and transparency particularly relevant.
The choice made here is clear: better to have something slightly less powerful but fully explainable. Not because speed doesn't matter, but because trust and care come first. In the end, the ability to justify decisions to a customer or a court is what determines the value of the solution.
Not working faster, but deciding smarter
Although AI is often positioned as an accelerator, Veenhof sees the greatest value lying elsewhere. Not in speed, but in the quality of the decisions being made.
Better decision-making eventually leads to more efficient processes on its own. But when speed is the starting point without attention to quality, there's a risk that mistakes just get made faster. The distinction isn't in what AI can do, but in how organisations use it.
"The real value of AI isn't in speed — it's in better choices. And those are what end up making everything faster."
The question every executive needs to ask themselves
By the end of the conversation, one core insight remains. Applying AI successfully doesn't start with technology, but with a simple, sharp question:
Is this actually on the boardroom agenda, and is it part of our strategy?
That brings AI back to where it belongs: not as hype or experiment, but as an integral part of how organisations create and deliver value. Organisations that get this right don't stand out because of the tools they use, but because of the choices they make.