India’s Debate Over Granting Legal Personhood to Artificial Intelligence Agents Raises Questions of Corporate Accountability

The Ministry of Corporate Affairs, responding to a series of parliamentary inquiries, has announced a draft amendment to the Companies Act 2013 that would ostensibly permit artificial intelligence-driven software entities to be recognised as legal persons for the limited purpose of entering contracts, owning assets, and being sued or sued by other parties. The proposal, presented amid heightened public concern over the opacity of algorithmic decision‑making, has ignited a vigorous discourse among jurists, economists, and technologists regarding the suitability of extending centuries‑old corporate doctrines to entities that lack consciousness, motive, or the capacity for moral judgement.

In the preceding twelve months, at least three independent financial technology firms have deployed autonomous advisory algorithms that, operating without transparent human oversight, executed high‑frequency trades which collectively resulted in an estimated loss of over two hundred crore rupees for retail investors, thereby prompting the Securities and Exchange Board of India to issue a formal warning regarding the absence of clear accountability in such cases. The Reserve Bank of India, citing similar irregularities in algorithmic credit‑scoring platforms, has further indicated that the existing framework of instrument‑based liability may prove insufficient to compel remedial action when the offending entity is a self‑learning codebase lacking a natural person to bear responsibility.

Legal scholars, invoking the precedent of corporate personhood established in Salomon v. A. Salomon & Co. Ltd., argue that extending the same doctrinal reasoning to non‑organic intelligences could erode the very basis of fiduciary duty, given that an artificial agent cannot possess intention, conscience, or the capacity to act in good faith, yet may nonetheless reap the benefits of limited liability. The draft bill, however, attempts to mitigate such concerns by proposing a statutory “electronic director” mechanism, whereby a designated natural‑person fiduciary would be required to certify the compliance of the AI entity with all applicable statutes, taxation obligations, and consumer‑protection regulations, thereby creating a hybrid responsibility structure that has yet to be tested in Indian jurisprudence.

Critics warn that the envisaged sanctions, which include monetary penalties proportionate to the AI’s aggregate turnover and the potential for compulsory dissolution of the digital entity, may prove symbolic at best unless coupled with robust enforcement powers granted to the Ministry of Corporate Affairs and the Securities and Exchange Board to impose personal liability upon the human stewards who ultimately wield the code. Moreover, the proposed amendment stipulates that any breach resulting in demonstrable consumer harm shall trigger an automatic suspension of the AI’s legal capacity to transact, a provision that, while ostensibly protective, raises practical questions about the technical feasibility of halting distributed ledger operations that may have already propagated across international networks.

From the standpoint of labour economics, the prospect of granting artificial agents quasi‑juridical status may accelerate the displacement of junior compliance officers, data analysts, and junior auditors, a development that could aggravate existing unemployment pressures in urban centres whilst simultaneously concentrating decision‑making authority within a small cadre of technocratic executives able to command sophisticated algorithmic platforms. Consumers, meanwhile, are left to navigate an increasingly opaque marketplace where the promise of efficiency is often juxtaposed against the spectre of anonymity, thereby undermining the traditional mechanisms of redress that rely upon the identification of a responsible corporate officer or board member.

If the legislature proceeds with bestowing legal personhood upon artificial intelligence agents, what safeguards will be instituted to ensure that the prescribed electronic director cannot merely become a nominal figurehead, thereby allowing the underlying code to evade substantive oversight and rendering the statutory liability regime a mere façade? Furthermore, in the event that an autonomous system incurs liabilities exceeding its declared capital, should the State reserve the prerogative to levy against the underlying shareholders, the appointed human fiduciary, or to invoke a newly created public fund financed by levies on all AI‑enabled enterprises, and what precedent would such a decision set for future corporate governance reforms?

Considering the cross‑border nature of many algorithmic platforms, how will Indian regulators coordinate with foreign jurisdictions to enforce suspension orders or to recover damages when the offending AI entity operates on servers situated beyond national territory, and does the proposed framework adequately address the complexities of extraterritorial jurisdiction and mutual legal assistance? Lastly, should empirical evidence emerge that consumer grievances are disproportionately directed at AI‑driven services, will the Parliament contemplate revisiting the foundational principle that only natural persons may bear criminal responsibility, thereby confronting the philosophical dilemma of attributing culpability to an entity devoid of consciousness, and what legislative mechanisms might be devised to reconcile this tension without eroding the rule of law?

Published: June 7, 2026