Algorithmic Mediation of Strategic Judgment: Executive Attention, Decision Architecture, and Power Reconfiguration under Generative AI

Abstract

Strategic management theory has traditionally conceptualized executive strategic judgment as a human-centered process shaped by managerial cognition, attention allocation, and hierarchical authority. The growing integration of generative AI into executive workflows challenges this view by introducing an algorithmic layer that structures how issues are surfaced, options are framed, and decisions are legitimated. This article develops a mechanism-based framework explaining how generative AI mediates executive strategic judgment through four processes: algorithmic attention structuring, probabilistic framing of strategic options, recalibration of epistemic authority, and temporal compression of decision cycles. Integrating the attention-based view of the firm, upper echelons theory, behavioral strategy, and research on algorithmic decision structures, the analysis shows that AI centrality generates non-linear effects on long-term strategic coherence. At moderate levels, algorithmic mediation enhances informational integration and responsiveness; beyond a threshold, convergence of salience, reduced framing diversity, authority centralization, and temporal compression may undermine exploratory capacity and long-term adaptability. Governance architecture, executive cognitive heterogeneity, environmental volatility, and organizational slack emerge as key boundary conditions. By repositioning generative AI as a constitutive element of executive decision architecture rather than a performance tool, this study advances a revised understanding of the microfoundations of corporate strategy in digitally intensive environments.