Based on: “2026 Work Trend Index: Agents, human agency, and the opportunity for every organization” (Microsoft, 2026)
For two years, most of us used AI to answer questions. Summarize this. Draft that. Useful, but still just a clever assistant waiting to be asked. Now agentic AI is changing that.
Microsoft’s latest Work Trend Index surveyed 20,000 people who use AI at work across ten countries, including the Netherlands. The finding is clear. AI is moving from answering questions to doing the work itself. These tools no longer just respond. They act. It sounds like a small change in wording. In practice, it is a big one.
From asking to delegating: the agentic AI ladder
The report describes four ways people work with AI. Together, they form a ladder.
First, you are the author. You do the work and call on AI here and there. Next, you become the editor. You set the direction, and AI drafts while you approve. Then comes the director. You hand off a whole task, and AI runs it in the background. Finally, you are the orchestrator. You design a system where several agents work in parallel, and you step in only for the exceptions.
The first two rungs are about answering. The last two are about doing. As organizations climb, the human role does not vanish. Instead, it changes shape. You spend less time on step-by-step execution. You spend more on setting direction, defining the standard, and judging the result. So the better AI gets at the work, the more your judgment matters, not less.
The real bottleneck is not the technology
Here is the finding every leader should sit with. The biggest barrier to value from AI is not the technology. It is not the people either. Rather, it is the organization around them. In fact, Microsoft attributes about two-thirds of AI’s real-world impact to organizational factors. Think culture, management, and the way work is structured. Only a third comes down to individual skill. In other words, people are ready. Most organizations are not.
Anyone who read our earlier piece on data quality will recognize the pattern. The technology is rarely the hard part. The foundation underneath it is. As AI shifts from advising to acting, three things stop being optional: the quality of your data, the clarity of your processes, and the guardrails that keep it accountable. Get them right, and agentic AI becomes leverage. Skip them, and it becomes a liability.
The Dutch Caribbean reality
The momentum here is real, and it is local. In May, the University of Curaçao led an AI knowledge mission to Silicon Valley. Our regulators and Social and Economic Council are already turning those lessons into a national strategy, with financial services in focus.
However, the real gap shows up in conversation. Over the past weeks, I have spoken with friends, relatives, clients, and prospects, both here and in the US. The same pattern keeps surfacing. Many organizations have rolled out an approved AI tool, but it underwhelms in daily use. So people quietly pay for a better personal subscription. Then they use it for work, without IT ever knowing. This is shadow AI. In an economy built on confidentiality and compliance, it is a real exposure. Moreover, it grows sharper as AI starts to act. An unsanctioned chatbot is one risk. An unsanctioned agent working on your data is another.
Banning these tools only drives the behavior deeper underground. The better answer is to bring it into the open.
What this means for you
- Fix your data foundation first. An agent is only as reliable as the data it acts on.
- Decide where AI should act, and where a human stays in the loop. Not every task belongs at the top of the ladder.
- Assume shadow AI is already happening. Meet it with sanctioned tools that are genuinely good, not with a ban that pushes it underground.
- Put the rules in writing. Create a clear AI usage policy. Name the approved tools and the prohibited ones. State what data may be shared, and what may not. Then train your people on the risks, so the policy is easy to follow.
Conclusion
The winners with agentic AI will not be the fastest adopters. They will be the most deliberate ones.
Be ambitious about the destination. Be deliberate about the path.
You can explore Microsoft’s full 2026 Work Trend Index here.
