ABSTRACT
Historically, corporate jurisprudence has relied upon legal fictions to attribute the actions of an artificial entity to an identifiable human mind. The increasing integration of autonomous algorithmic systems into corporate strategy places significant strain on this architecture. This paper examines the doctrinal vacuum created by Artificial Intelligence within Indian corporate governance, specifically under the Companies Act, 2013. When boards delegate commercially consequential decisions to opaque machine-learning models, traditional frameworks governing fiduciary duties and corporate attribution become exceedingly difficult to apply coherently.
This article demonstrates how reliance on algorithmic outputs conflicts with the non-delegable duties of independent judgment and due care mandated by Section 166. Furthermore, it argues that AI deployment complicates the continued application of the alter ego doctrine, potentially insulating corporate officers from intent-based criminal liability and vicarious civil claims. While emerging regulatory trajectories—such as the European Union’s horizontal risk categorization and India’s sector-specific guidelines—address product safety and market integrity, they largely bypass this internal fiduciary crisis.
To resolve the resulting liability vacuum, the paper proposes a targeted evolution of Indian corporate law. Rather than artificially stalling technological adoption, the analysis suggests that statutory frameworks must continue to attach algorithmic deployment to meaningful human oversight. The paper recommends specific legislative interventions, including updating directors’ duties to mandate algorithmic diligence, expanding disclosure requirements under Section 134 to require independent algorithmic audits, and cautiously narrowing the judicial protection of the business judgment rule. Ultimately, the research concludes that while AI will inevitably reshape commercial operations, corporate law must ensure that autonomous technological systems do not gradually dilute the foundational premise of human accountability.
KEYWORDS
- Corporate Governance
- Artificial Intelligence
- Fiduciary Duties
- Corporate Attribution
- Algorithmic Accountability
- Business Judgment Rule
- Companies Act, 2013
PART I: INTRODUCTION
Artificial Intelligence is no longer a futuristic concern for corporate governance in India; it is already embedded within day-to-day boardroom decision-making. Large corporations increasingly rely upon algorithmic systems to execute high-frequency trading, assess merger and acquisition opportunities, predict market behavior, and streamline recruitment processes. On Indian stock exchanges alone, algorithmic trading now constitutes a substantial share of total trading activity.[1]What was once considered an experimental technological tool has gradually become part of the operational structure of modern corporations. Decision-making authority, historically exercised by directors and senior officers, is increasingly being delegated—sometimes almost invisibly—to autonomous software systems capable of acting at speeds and scales beyond ordinary human supervision.
This development creates a serious tension within traditional corporate law. Modern company jurisprudence was constructed on the assumption that corporate conduct ultimately flows from human judgment. Concepts such as the “directing mind and will” doctrine, fiduciary obligations, and corporate criminal liability all presuppose the existence of identifiable human intention.[2] Even the doctrine of mens rea, as applied to corporations, depends upon attributing a guilty state of mind to natural persons acting on behalf of the company. Artificial Intelligence complicates these assumptions in ways existing legal doctrine struggles to accommodate. An algorithm may influence or even independently produce decisions with significant economic consequences, yet it possesses neither legal personality nor moral agency. If an autonomous trading system manipulates markets, or an AI-driven hiring tool systematically discriminates against applicants, the immediate question becomes difficult to answer: where exactly should the law place responsibility?
The uncertainty is especially visible within the Indian regulatory framework. Existing statutes, including the Companies Act, 2013 and the Information Technology Act, 2000, were enacted at a time when autonomous corporate decision-making was not a realistic legislative concern. Section 166 of the Companies Act imposes duties of good faith, due care, and independent judgment upon directors.[3] However, the legislation provides little guidance on how those duties are to be discharged when directors rely on opaque “black box” AI systems whose reasoning processes remain inaccessible even to technical experts. The problem becomes more acute where the consequences are severe. If an AI model causes substantial shareholder losses, facilitates anti-competitive conduct, or produces discriminatory corporate outcomes, Indian law presently offers no coherent framework for apportioning liability between directors, corporations, software developers, and third-party vendors. The result is a regulatory vacuum in which technologically significant decisions may occur without corresponding legal accountability.
This paper argues that corporate law can no longer remain anchored exclusively to a model of pure human agency. Instead, the legal framework governing corporate decision-making must evolve toward a principle of human accountability for algorithmic actions. Corporations should not be permitted to avoid liability merely because critical decisions were generated through automated systems. Accordingly, this paper contends that Indian corporate regulation requires mandatory algorithmic audits, enhanced disclosure obligations relating to AI-assisted decision-making, and a clearly enforceable duty of algorithmic oversight imposed upon boards of directors. Without such reforms, the growing integration of Artificial Intelligence into corporate governance risks creating a system in which economically consequential decisions are made through mechanisms that remain legally under-regulated and practically unaccountable.
The discussion proceeds in five parts. Part II examines the manner in which AI systems disrupt traditional understandings of directors’ fiduciary duties and the “directing mind and will” doctrine. Part III addresses the emerging liability problem, particularly in relation to negligence, corporate criminal liability, and shareholder remedies where algorithmic systems malfunction or act unpredictably. Part IV undertakes a comparative analysis of India’s current position alongside evolving international approaches, particularly the European Union Artificial Intelligence Act. Finally, Part V proposes specific statutory and regulatory reforms aimed at integrating accountability, transparency, and algorithmic oversight into the Indian corporate governance framework.
[1]Securities and Exchange Board of India, Consultation Paper on Guidelines for Responsible Usage of AI/ML in Indian Securities Markets (June 20, 2025).
Lennard’s Carrying Co Ltd v Asiatic Petroleum Co Ltd [1915] AC 705 (HL) (establishing the ‘directing mind and will’ doctrine, attributing corporate actions to senior individuals).
[3]Companies Act, 2013, § 166(2)–(3).