AI as Strategist

Artificial intelligence (AI) systems are increasingly touted as potential substitutes for human strategists. Building on van den Steen’s formal theory of strategy, this paper develops a model in which the strategist may be a human decision maker or an AI. The model characterises how differences in information processing, belief accuracy and credibility affect a strategist’s ability to guide interdependent choices through influence or formal control. We show that strategic value arises through two channels: a direct improvement in the quality of individual decisions and a coordination effect that aligns decisions with one another. While AI exhibits superior analytical capabilities in structured, data-rich environments, its ability to persuade and coordinate depends critically on its transparency and the degree of agreement it can generate with human decision makers. As a result, AI requires less formal authority precisely where its analytical advantages are greatest, whereas human strategists retain advantages in judgment-rich contexts. These findings imply that organisations should adopt domain-contingent approaches to integrating AI into strategic processes, tailoring the allocation of control and influence to the nature of each decision. This framework extends strategy theory to technological agents and offers practical guidance for managing human–AI collaboration in strategy.