How bold leadership with AI makes the difference between yet another project and a real leap in performance
AI is now on virtually every boardroom agenda.
Pilots have been launched, task forces set up, innovation hubs opened. Consultants present opportunities, and vendors show impressive demos. In presentations and reports, the organization seems to be in full swing.
But when you zoom in on what customers and employees actually notice, the picture is often much less impressive. Turnaround times barely change. Errors recur stubbornly. Workloads on teams remain high. The promise of AI – smarter, faster, better – still rarely translates into concrete noticeable improvement.
That’s not a shortage of technology.
That’s mostly a shortage of direction.
This article is part of an inspiration campaign for directors, as part of the executive leadership program – Innovation & Transformation with Data and AI, developed in collaboration with the Erasmus Centre for Data Analytics, part of Erasmus University Rotterdam. The central message: AI is not a techno thingy, but a matter of vision, strategy, choices and leadership in the boardroom.
When AI is an agenda item, but not a decision
The picture is often recognizable. Late in the afternoon there is an agenda item “AI – state of play” in the board meeting. The CIO or CDO gives an update: several pilots are running, there is a proof of concept with generative AI, there are talks with a few major suppliers. There is a nod, an occasional critical question, appreciation for the energy. The conclusion at the end is usually, “Keep going, keep us informed.”
What is not spoken is just as important:
- where AI in this organization ultimately MUST make a difference for customers and citizens;
- What specifically may change in the daily work of employees;
- What limits there are from values, responsibility and risk.
This creates a situation where “something is happening with AI” all over the organization, but little is explicitly decided in the boardroom. AI becomes a movement without ownership. It arises, rather than being chosen.
The question that then remains under the table is sharp but simple:
- If we as a board do not lay down a clear vision for AI, are we willing to accept that others will implicitly fill it in for us?
AI as a strategic choice, not a tool
The only truly defensible reason to take AI seriously in the boardroom is that it contributes to something that is strategically crucial to you: the next level of operational excellence, with customers and employees at the center.
This becomes concrete when you couple it with three observations:
- Customers still accept too much hassle: unclear lead times, varying responses, lots of “behind-the-scenes” hullabaloo.
- Employees deliver top performance every day despite systems, not because of them. They search, correct and record their way through the day.
- The operation as a whole could arguably be smarter, but a true breakthrough level is rarely achieved.
It is precisely in that area of tension that AI has meaning. Not as a loose tool, but as a strategic choice: we deploy this technology where the gap between potential and performance is greatest. Where better information, faster signals and less repetitive work are directly felt by customers and employees.
Without that linkage to the core of the operation, AI becomes a sideline: an experimentation angle, an innovation plaything. With that linkage, AI becomes a tool to structurally increase performance.
- If you were allowed to deploy AI in only one place in your organization, where the gap between potential and performance is greatest right now, where un your organsation would you specifically choose and why exactly there?
Without a vision, AI mostly accelerates the old pattern
Technology rarely breaks existing patterns by itself. More often, it reinforces what is already there. An organization with a fragmented change agenda will get more projects, not less, thanks to AI. An organization with too many priorities will gain new ambitions, but not additional focus. An organization that is project weary quickly experiences AI as “the next program we have to go through.”
A clear vision of AI is therefore not a formality, but a necessary condition.
Such a vision not only describes what is technically possible, but especially why and to what purpose you are using AI. It makes it explicit:
- Which customer and operations issues you prioritize;
- What values you hold (people-centeredness, explainability, fairness, data security);
- what the division of roles between humans and AI looks like: what AI may prepare and signal, where humans remain unabated.
Lacking that vision, the debate shifts to the layers below the boardroom. IT, innovation and business try to decide among themselves “what to do with AI.” The result is delay, confusion and an AI portfolio that no one can really explain.
With a sharp vision, each proposal becomes easier to assess: does it demonstrably contribute to what we have expressed – or not?
- If you look honestly at the current AI initiatives in your organization: do they primarily increase the number of projects or do they demonstrably bring you closer to a sharply articulated governance vision of what AI is and is not for?
The silent condition: do you dare to trust your data?
AI and data are inextricably linked. Yet in practice, data quality often remains a technical topic, far away from the boardroom. While that is precisely where the greatest tension lies.
AI puts a magnifying glass on the quality of your data. Models that train on incomplete, inconsistent or outdated data produce answers, but not necessarily reliable ones. Reports look impressive, dashboards sleek – while under the hood the basics are shaky.
Then AI suddenly becomes a question of legitimacy:
can we still convincingly explain the decisions that arise from it to customers, regulators or media?
The question directors must then ask themselves is uncomfortable, but essential:
- On what data do we really dare to base decisions that could be in the newspaper tomorrow?
If the answer is hesitant, there is a data issue first.
And that is not an IT issue, but a strategic issue.
The real bottleneck: not technology, but learning capacity
Among executives we speak to, the picture is remarkably consistent: technology is developing rapidly, but their own organizations are learning too slowly. Not because people don’t want to, but because rhythm and structure are lacking.
Operational excellence in an AI era is all about clock speed: how often and how well an organization learns from data, from customers and from employees. AI can greatly increase that clock speed, but only if there are decision moments and learning loops in which those insights are actually used.
Without that rhythm, AI becomes a mirror no one looks into. With that rhythm, AI becomes an accelerator of everything you understand by good governance and excellent execution.
This article is an invitation to no longer see AI as something that “happens somewhere in the organization,” but as a touchstone for leadership in the boardroom. The inspiration campaign and leadership program we are developing with the Erasmus Centre for Data Analytics, part of Erasmus University Rotterdam, focuses precisely on this shift.
The question is not: what can we all do with AI?
The question is:
- What improvement do you think is so crucial
that you dare make AI indispensable for it –
and what vision do you lay down for it in your own boardroom?
