Abstract
Comparative research on small and medium-sized enterprises (SMEs) and artificial intelligence usually treats national institutions and firm leadership separately. Institutional studies ask which countries make adoption easier. Leadership studies ask which managers make adoption happen. In founder-led SMEs, where one person often holds most decision rights, these questions cannot be separated. This study argues that the same founder can be more or less effective at leading generative AI adoption depending on the institutional setting, and that the direction of this effect depends on the type of readiness the founder has. We develop the Founder-Institution Fit (FIF) model. It builds on the Founder Readiness Framework (FRF-4) proposed by Sarkisyan (2026) and combines two founder composites, structural and psychological readiness, with three national parameters: the depth of the market for external expertise, the strength of internal delegation norms, and regulatory clarity. We apply the model to four stylized founder archetypes in Switzerland and the United Kingdom. Both are high-income economies, but they differ sharply in institutional type. The illustrative analysis gives three findings. First, psychological readiness works as a multiplier. A founder who is low on it achieves little in either country, whatever institutional support is available. Second, the country advantage reverses with the founder's profile. Founders whose readiness rests mainly on learning orientation gain more in the British setting, which offers a deep market for outside expertise. Founders whose readiness rests mainly on identity plasticity gain more in the Swiss setting, which supports internal delegation. Third, once regulatory differences are neutralized, the crossover follows a closed-form threshold: the British setting is preferred when the ratio of learning orientation to identity plasticity exceeds 4:3. We discuss implications for founders, boards, advisors, and policymakers and set out an agenda for empirical calibration.