Business strategy for an AI-shaped market
The strategic questions that have changed — and the ones that haven't.
Some strategic questions have changed. Many have not, and the confusion between the two is doing damage.
What has changed: the unit economics of knowledge work, the speed at which a competitor can replicate a feature, the shape of the buyer journey when agents are part of it, and the geopolitical layer around where models and data sit. Switzerland's position between US and Chinese AI ecosystems is now a strategic variable, not a footnote. Digital sovereignty is a procurement line, not a conference theme.
What has not changed: whether you have a defensible position, whether your customers would notice if you disappeared, whether your organisation can actually execute what the strategy document promises, and whether your capital structure matches the time horizon of the bets you are making. AI does not fix a weak position. It accelerates whatever position you have, in whichever direction you were heading.
The strategic error we see most often is treating AI as the strategy. It is not. It is a capability layer that reshapes the economics of several existing strategic questions. The questions themselves — who you serve, how you win, what you refuse to do — are the same ones they have always been.
The firms that remember this will spend 2026 making better decisions. The ones that forget it will spend 2026 writing AI strategies that read well and change little.
The decision: use AI to sharpen the strategic questions you already had — not to replace them with new ones you do not.