Abstract
The swift deployment of Artificial Intelligence (AI) enabled solutions into business decision-making processes has revolutionised the business operations within modern enterprises. Automated financial assessments and compliance and risk management systems powered by algorithms are emerging as corporate systems that depend more on algorithmic decision-making to enhance efficiency and accuracy. Yet this move into technology creates complicated jurisprudential issues around liability and accountability, especially when such systems create wrong, prejudiced or detrimental conclusions. This section analyses the issues presented by algorithmic decision-making in corporate India, specifically, how the existence of the existing legal infrastructure fails to address responsibility and attribution of fault. Traditional corporate governance processes are based on human decision-making, and, the increasing dependence on autonomous or semi-autonomous AI makes it difficult for liability of corporate entities, directors and technology developers to be identified. Taking a doctrinal and analytical approach, it discusses what AI will become, how it is emerging as a player in corporate governance and in particular, the implications of both relevant laws (Companies Act) and regulatory norms in relation to that in corporate behaviour. And it does a comparative reading of the developing global regulatory initiatives specifically on regulatory regimes that differ in the European Union to identify some reform possibilities. The paper asserts that the lack of a coherent legal system governing AI-based corporate action poses wide gaps to regulation, and hence the need to re-evaluate accountabilities under existing ones. Lastly, it suggests an appropriate legal and governance framework designed to improve transparency, fairness and liability definition pertaining to algorithmic corporate governance in India.
Keywords: Artificial Intelligence, Algorithmic Decision-Making, Corporate Governance, Accountability and Liability, Regulatory Framework.
Introduction
The increased deployment of Artificial Intelligence (AI)[1] in corporate operations has significantly changed the nature of decision making in today’s business ecosystem. Algorithmic systems[2] enable corporates to perform many functions todayfinancial analysis, risk assessment, regulatory compliance, recruitment and strategic planning. These systemsempowered by complex data-processing capabilitiesare often seen as more efficient and objective,[3] capable of minimizing human error.
Yet a reliance on algorithmic decision-making prompts important legal and ethical questions[4] too, especially when such systems yield erroneous, discriminatory, or harmful results. In the classical company model, accountability and liability are vested in identifiable human actors such as the director, the officer or the employee whose decisions can be examined in a legal framework. The deployment of AI decision-making subverts this system by introducing aspects of autonomy as well as opacity, a concept known as the ‘black box’ dilemma,[5] where the rationale for what is generated based on the algorithmic results is not so easily understood. This change threatens the pillars of corporate governance[6] in the sense of transparent, accountable and ethical management.
The Indian legal regime including Companies Act[7] and other supporting framework of law and regulatory mechanism has not really yet caught up with new challenges associated with the algorithmic system in corporate decision making. Though these legislations give an organized structure to what is necessary to govern corporations legally and legally responsible for liability, they are still the root of human agency,[8] and they never quite capture all the time when AI technologies shape and or set decisions. Thus, when algorithmic decisions cause loss, non-compliance, or violation of stakeholders, a considerable loophole widens in determining liability.
This problem becomes urgent following rise of AI technologies for Indian corporates as they are undergoing digital transformation[9] and being up against the global competitive pressures. Without clearly defined legal principles, this begs the question of whether liability rests with the enterprise, the board of directors, or the developers[10] and designers of the algorithmic systems. Furthermore, concerns over bias,[11] lack of disclosure and responsible conduct make it even more difficult for the regulatory environment.
The objective of this paper is to explore the legislative question-challenges related to algorithmic decision-making in the Indian corporate sector. Its intention is to analyse if the existing corporate governance and liability provisions are capable to deal with such challenges, whilst establishing which loopholes should be targeted for legal change.In adopting a comparative point of view and analysing the international regulatory approach,[12] the research tries to put forward one specific framework, making AI-based investment in the corporate world accountable, transparent and just.
[1] Stuart Russell & Peter Norvig, Artificial Intelligence: A Modern Approach (4th ed., Pearson, 2021).
[2] Tarleton Gillespie, “The Relevance of Algorithms” in Media Technologies (MIT Press, 2014).
[3] Cary Coglianese & David Lehr, “Regulating by Robot: Administrative Decision Making in the Machine-Learning Era” (2017) 105 Georgetown Law Journal.
[4] Luciano Floridi and others, ‘AI4People—An Ethical Framework for a Good AI Society: Opportunities, Risks, Principles, and Recommendations’ (2018) Minds and Machines.
[5] Frank Pasquale, The Black Box Society: The Secret Algorithms That Control Money and Information (Harvard University Press 2015).
[6] OECD, G20/OECD Principles of Corporate Governance (OECD Publishing 2015).
[7] The Companies Act 2013.
[8] Julia Black, ‘Decentring Regulation: Understanding the Role of Regulation and Self-Regulation in a “Post-Regulatory” World’ (2001) 54 Current Legal Problems.
[9] NITI Aayog, National Strategy for Artificial Intelligence (2018).
[10] Ryan Abbott, ‘The Reasonable Robot: Artificial Intelligence and the Law’ (2018) 105 Georgetown Law Journal.
[11] Solon Barocas and Andrew D Selbst, ‘Big Data’s Disparate Impact’ (2016) 104 California Law Review.
[12] European Union, Artificial Intelligence Act (2024).