Strategy in an Exponential AI World

Following on from my last piece, which explored the difference between transformation and optimisation, I have found myself reflecting on another word that appears repeatedly in discussions about change, notably ‘strategy‘. It is a word used with confidence and frequency, often as the organising discipline for transformation. Yet in an environment shaped by exponential AI driven change, it is worth asking whether our traditional approach to strategy is still sufficient or whether it risks becoming a way of refining yesterday’s assumptions rather than preparing for an uncertain future.

Traditional strategic planning assumes that change occurs at a pace that can be reasonably forecast, modelled and managed through linear roadmaps, annual budgets and multi-year plans. However, the emergence of frontier AI, autonomous agents, synthetic content, accelerated software development and increasingly adaptive cyber threats is compressing technology cycles and creating levels of uncertainty that traditional planning methods were never designed to accommodate. In an AI augmented world, capability curves can move faster than planning cycles, competitors can emerge from unexpected places and assumptions can expire before the strategy is fully implemented.

This creates a challenge for traditional strategic planning. That approach is useful when change is incremental. Traditional financial forecasting suffers from the same weakness. Forecasting is useful for budgeting, accountability and operational control but it is usually anchored in historic performance. In a world of exponential AI driven change, it can look precise while being strategically misleading. It risks turning the future into an extrapolation of the past. The concern should be not that the business is standing still but that it may be evolving from where it has come from, rather than transforming for the realities of that uncertain future.

In this environment, the challenge is no longer simply predicting the most likely future but preparing for multiple plausible futures that may emerge far more rapidly than expected.

Scenario planning provides a more resilient approach, a better strategic lens, by exploring a range of credible technological, regulatory, geopolitical and threat driven outcomes, enabling organisations to identify strategic options, stress-test assumptions, understand vulnerabilities and build adaptive capabilities before disruption occurs.  It does not claim to predict what will happen. Instead, it explores what could happen. It forces leaders to confront uncertainty, discontinuity and second order effects such as rapid AI commoditisation, regulatory constraint, labour disruption, new entrants, radical productivity gains or shifts in customer expectations.

The point is not to abandon planning or forecasting. It is to reposition them. Forecasts manage the current business. Traditional plans coordinate execution. Scenarios prepare the organisation for multiple futures.

Rather than seeking certainty in an increasingly uncertain world, scenario planning equips leaders to make informed decisions under uncertainty, improving organisational resilience, security and competitiveness as AI continues to reshape both the opportunities and risks facing modern enterprises.

In an exponential AI world, the winners are unlikely to be those with the most elegant 5 year plan. They will be those with clear intent, adaptive capacity, evolved strategic thinking and the courage to transform before optimisation becomes a trap. Traditional forecasts will remain useful for managing today’s business but they should not be mistaken for a reliable map of tomorrow’s. As AI commoditises todays capability at source, the financial logic of the past is very likely going to bear little resemblance to the economics of the future. Ask yourself, are you optimising for a world that is already disappearing?