A year ago, our conversations with Belgian leaders were mostly about where to start with AI. Then the focus moved to getting people to use it. More recently, the conversation has changed again: is the investment in AI actually paying off?
The KPMG Global AI Pulse Q3 2026 suggests that organizations are getting better at answering that question. Eighty-nine percent now report measurable value from AI. Yet only 12 percent consistently compare that value with what their AI costs to run. As AI scales, the question is increasingly about which investments are worth scaling and which are not.
That gap is telling. The challenge is shifting from getting AI adopted to making conscious choices about where it creates real value, what is worth scaling, and how to manage it once the business starts to depend on it.
We see that shift in Belgium too. Cost is increasingly becoming part of the management conversation. Sovereignty, which until recently could feel like a technical topic, is also becoming much more concrete: how dependent are we on particular models or providers, and what happens if that dependency changes? The global survey reinforces that shift: 72 percent of respondents now give model sovereignty a formal role in decision-making.
These are quite different questions from the ones we were discussing a year ago. It is no longer just about getting AI adopted. It is about making deliberate choices about where it creates value, what it costs, and how much we want to depend on it.
KPMG’s Global AI Pulse Q3 2026, based on a survey of more than 2,100 senior leaders across 20 countries, territories, and jurisdictions, highlights how competitive advantage is moving from simply adopting AI to running it effectively.