When people ask me about leadership, they often expect a story about vision, inspiration, big transformation arcs. I understand why. It is the version of leadership that travels well.
My day-to-day reality is simpler, and tougher. Most of the time, I am not trying to be inspiring. I am trying to keep the organisation coherent while several forces pull it in different directions at the same time.
AI is now an expectation, not a curiosity. Hybrid cloud is no longer a transitional phase, it is the baseline most enterprises live in. Security standards keep rising, regulations keep tightening, budgets remain under pressure, and talent is still difficult to find in the exact places you need it. Add the fact that an international group has strong local leadership and strong local realities, and you quickly understand why leadership in IT is rarely a clean sequence of decisions.
I spend a lot of time on trade-offs. Not because I enjoy complexity for its own sake, but because trade-offs are what the job is made of.
There is the trade-off between speed and control. There is the trade-off between local autonomy and group coherence. There is the trade-off between investing in the future and protecting today’s operational stability. There is the trade-off between internal capability and external acceleration. Most of these trade-offs do not resolve themselves. They require explicit decisions, and they require me to explain the rationale behind them in a way that teams can actually work with.
Hybrid cloud is a good example because it is one of those topics where the slogan version of the conversation is useless. In reality, hybrid means distributed responsibility. Data sits in different places. Workloads run across different environments. Vendors change their terms. Security models are layered. Costs are dynamic. The complexity is structural. It does not disappear because someone publishes a framework. If I want the organisation to operate safely and efficiently in that environment, my job is to create an operating model that holds together cost visibility, security posture, data governance, and delivery speed without turning everything into a bureaucratic maze.
AI adds another layer of pressure, and I think 2026 is the year when the tone changed. The question is no longer whether AI matters. The question is where it creates measurable value and where it creates noise, risk, or distraction. I see the same pattern everywhere. People love pilots because pilots feel fast. Industrialising AI is slower because it forces uncomfortable work on data quality, governance, security, and change management. If I do not protect teams from chasing everything, we end up with a portfolio of half-finished experiments and a credibility problem.
Leadership in this context is mostly discipline. It is deciding what we will scale, what we will stop, and what we will not start yet. It is also building internal capability so that we do not outsource our future by accident. If the organisation wants AI to be a real capability, then my teams must be able to operate it, govern it, improve it, and explain it. That does not happen through announcements. It happens through deliberate investment in skills, in architecture, and in operational routines.
What makes this harder in a Group environment is that my decisions land in different realities. A regional CIO will look at speed, local constraints, client expectations, and the operational impact of a standard. Group leadership will look at risk, cost, resilience, and consistency. Both perspectives are rational. The job is to design standards that protect the backbone and leave enough flexibility for local execution to remain effective.
I have learned to be very explicit about what must be common and what can remain local. Identity and security baselines cannot vary by geography. Core data definitions cannot drift without consequences. Some infrastructure choices must be coherent if we want interoperability. Outside of that, there are areas where local adaptation is legitimate, and pretending otherwise usually creates workarounds that are worse than the variation we were trying to remove.
None of this works without communication that is steady and repeatable. I do not mean constant messaging. I mean the ability to explain the same direction consistently, in plain language, over time. People do not need perfect certainty from leadership. They need coherent reasoning. When decisions change, they need to understand why. When budgets tighten, they need to understand what is protected and what is deferred. When we standardise, they need to understand what problem we are solving and what trade-offs were considered.
If I had to describe leadership in IT in 2026 in one honest line, it would be this. My job is to absorb complexity so the organisation does not fragment under it, translate that complexity into choices people can execute, and keep the whole thing tied to business outcomes that matter beyond IT. There is less theatre in that than in most leadership narratives. There is more repetition, more structure, more uncomfortable prioritisation. There is also a constant effort to work across functions, finance, security, operations, regions, so decisions don’t stay trapped inside the technology bubble.
But that is exactly the point. When it works, IT becomes a transversal force that gives the Group the means to match its ambitions: a backbone that makes growth, resilience, and execution possible.
