AI and Strategic Decisions: Facing the Incompleteness

Abstract This study examines the role of artificial intelligence (AI) systems—and the related legal requirements—in strategic corporate governance decisions, characterised by a high degree of incompleteness and, consequently, discretion. It illustrates the inadequacy of current European regulatory developments in the information technology field, primarily based on a product liability approach, to govern the complexities underlying the introduction of AI tools in the strategic choices made by directors. Given the core elements of corporate governance and its specific accountability needs, the study demonstrates how algorithmic decision-making cannot be an exclusive strategy for strategic decisions characterised by unpredictability and incompleteness. Therefore, the ultimate objective of the analysis is to underscore the persisting gap in legislation regarding the need to shift from a product-based and risk-minimisation approach in AI-driven decision-making to an integrated approach wherein the decision-making process is shaped by the complex, yet balanced, interaction between two distinct decision-making components, namely humans and the algorithm. To this end, the paper advocates for a stronger emphasis on human enhancement, algorithm ergonomics, and a legal strategy-driven design of AI.