Artificial Intelligence and Fiduciary Duties: Rethinking Board Liability for Algorithmic Decision-Making

The integration of Artificial Intelligence (AI) into corporate decision-making is transforming the traditional structure of corporate governance. From financial forecasting and risk assessment to recruitment, contractual evaluation, fraud detection and strategic planning, AI-driven systems increasingly influence decisions that were historically undertaken by directors and senior management. Although these technologies offer efficiency, predictive accuracy and cost advantages, their deployment raises fundamental questions concerning accountability, transparency, independent judgment and the fiduciary responsibilities of corporate directors. Traditional company law presupposes that directors exercise personal judgment, act in good faith and discharge their responsibilities with reasonable care, skill and diligence. However, algorithmic decision-making complicates this assumption because decisions may be generated through complex systems whose internal reasoning is difficult to interpret, whose training data may contain historical biases and whose outputs may change through continuous learning. The central question is whether directors can discharge their fiduciary obligations merely by relying on technologically sophisticated systems or whether such reliance creates a heightened responsibility to understand, supervise and monitor algorithmic processes. This paper critically examines the relationship between AI-driven corporate decision-making and directors’ fiduciary duties, particularly under Section 166 of the Companies Act, 2013. It analyses the emerging challenges of algorithmic opacity, automation bias, accountability gaps, data governance and the delegation of managerial judgment to technological systems. Through comparative examination of the United Kingdom, United States and European Union, the paper evaluates how existing corporate governance principles may be adapted to address AI-related risks. It argues that the adoption of AI does not dilute directors’ legal responsibilities; rather, it requires a more sophisticated interpretation of their duties of care, diligence, good faith and independent judgment. The paper proposes a risk-based framework of AI governance involving board-level oversight, human accountability, algorithmic audits, explainability, continuous monitoring and documented decision-making.