AI-Driven Decision-Making: Balancing Effectuation and Causation in Entrepreneurial Decision-Making Strategy

Sammanfattning

In the rapidly evolving landscape of entrepreneurship, Artificial Intelligence (AI) has emerged as a pivotal tool reshaping how founders navigate uncertainty and allocate resources. Grounded in Sarasvathy’s dual logics of causation (predictive, goal-oriented planning) and effectuation (adaptive, resource-driven experimentation), this study investigates whether AI reinforces entrepreneurs’ default decision-making style or catalyzes a shift toward the alternate logic. Drawing on a global survey of 103 startup founders and CEOs—predominantly in technology and healthcare sectors—our research integrates validated measures of baseline decision orientation (EffBase, CauBase) with novel scales capturing AI prompt engineering practices (EffPrompt, CauPrompt). Hierarchical moderated regressions reveal a pronounced reinforcement effect: entrepreneurs predisposed to effectual reasoning who engage in adaptive, iterative prompting exhibit a significant amplification of their heuristic, exploratory approach (β = 0.503, p < .001; ΔR² ≈ 12.7%). In contrast, causal entrepreneurs do not experience a comparable enhancement through rigid, goal-focused prompts; if anything, a marginally negative interaction (β = –0.178, p = .058) hints at a nascent “shift” toward more flexible reasoning. These findings underscore that AI’s impact is not neutral—its strategic value hinges on how prompts are crafted. By situating AI as both amplifier and boundary condition of entrepreneurial cognition, this study bridges effectuation theory with contemporary AI strategy research. Practically, our results advocate for tailored prompt-engineering training: adaptive techniques to deepen effectual strengths, and deliberate interface designs that invite exploration even for goal-driven users. We conclude with recommendations for longitudinal and cross-cultural investigations, call for analysis of actual AI‐prompt logs, and propose linking prompting styles to concrete innovation outcomes, thereby charting a research agenda at the intersection of AI and entrepreneurial decision-making.