Legal Challenges and Regulatory Responses
AUTHORED – KANISHKAA KUNDU, 4th YEAR, BBA. LLB (HONS.) at Sister Nivedita University, Kolkata, West Bengal
ABSTRACT:
There is an ongoing process of incorporating Artificial Intelligence into the board room decision-making process, risk assessment procedures, compliance monitoring, and stakeholder engagements in India. Despite the efficiency and accuracy that can be achieved in corporate governance through artificial intelligence technology, there are also significant challenges in integrating emerging technologies due to their implications for the well-established concepts of directors’ duties, fiduciary responsibilities, and reporting obligations under the Companies Act, 2013. This paper will critically evaluate the challenges posed by incorporating artificial intelligence technologies into Indian corporate governance, which revolves around human-oriented concepts such as the “directing mind and will” concept. In particular, the paper will critically analyze SEBI’s response in relation to AI/ML regulations, including the Consultation Paper on SEBI (Amendment of Rules and Regulations) Relating to AI/ML (Up to 2025), SEBI (Intermediaries) (Amendments) Regulations, 2025, Regulation 16C, and the Digital Personal Data Protection Act, 2023.
TO THE POINT:
Indian corporate boards have increasingly adopted AI-driven technology in risk assessment modelling, fraud detection, ESG ratings, and even drafting board resolutions. These developments raise an important question of governance: Who is Held Liable in Case of AI Error?
Indian Company Law has been framed on the understanding that the decision-making process involves a natural person who possesses sound judgment, attends meetings, and takes decisions “with due and reasonable care.” AI presents an interesting challenge in this regard since the use of artificial intelligence involves output which may be probabilistic, opaque, and hard to verify ex-post.
There being no horizontal law governing AI in India, the regulation of artificial intelligence in the corporate sector amounts to a series of interventions in the sector-specific spheres, primarily by SEBI and the Reserve Bank of India, upon the general duty-based framework established by the Companies Act, 2013. It is this situation of regulatory fragmentation which makes it uncertain for the board directors as to what extent of reliance on AI outputs is acceptable without being a dereliction of duty.
USE OF LEGAL JARGON:
Company Law 2013 provides for the duty of directors by stating that they must be honest, independent, use their independent judgment, and take care while performing their duties. The “business judgment rule” has not been codified explicitly under the Indian laws, but it has been applied by the Indian courts by providing protection to the directors against any liability when the decision was well-informed, free from conflicts of interest, and made in good faith. While using AI tools for governance purposes, the problem arises because the notion of independent judgment and use of a black box recommendation system cannot coexist.
Regulation 16C, introduced under the SEBI (Intermediaries) Regulations by way of an amendment in February 2025, mandates that all intermediaries shall be strictly liable for the AI/ML technologies used by them, either developed internally or outsourced from third party vendors, in terms of data integrity, correctness of output generated and compliance with law. The principle of “sole responsibility” renders any attempt to argue an outsourcing of algorithm-based functions as untenable. On the other hand, the Digital Personal Data Protection Act 2023 creates liabilities for ‘data fiduciaries who process personal data of shareholders, employees or consumers for corporate governance purposes and could attract a fine of up to 250 crore rupees.
THE PROOF:
The case of urgency can be found in SEBI’s own actions. SEBI’s guidelines for responsible use of artificial intelligence/machine learning issued in its consultation paper dated June 2025, “Guidelines for Responsible Usage of AI/ML in Indian Securities Markets,” propose a board approved AI Governance Framework based on six principles – ethics, accountability, transparency, auditability, data privacy, and fairness – for all regulated entities. These include internal approvals, algorithmic audits, model explainability, fall back plans, and regular reports on accuracy.
The growing trend towards holding directors personally liable in cases of corporate governance failure in India can be seen in various other areas of law in India that have nothing to do with financial regulation, such as taxation and environmental laws, which has shown that courts will look beyond the formal structure of the corporation in cases of governance failure and hold individuals personally accountable. In sum, both Indian regulators and the Indian judiciary have reached a consensus on one issue: intermediation through technology has not reduced accountability; it has simply shifted the onus to demonstrate it.
CASE LAWS:
- Official Liquidator v. P.A. Tendolkar, (1973) 43 Comp Cas 382 (SC)
It was established by the Supreme Court that directors failing to exercise reasonable supervision over management affairs can be held responsible for negligence and breach of duties, despite no involvement in the offending act. The ruling can be very important for the issue of AI governance because the directors’ inaction regarding the algorithmic recommendation can also result in liability.
- Miheer H. Mafatlal v. Mafatlal Industries Ltd., (1997) 1 SCC 579
The Court provided criteria very similar to the concept of business judgment rule, according to which courts cannot interfere in management’s decisions, and only if the board’s procedure was rational and made in good faith. Such standards would definitely play an important role in future cases concerning AI governance when it comes to examining the process of decision-making.
- N. Narayanan v. Adjudicating Officer, SEBI, (2013) 12 SCC 152
This case underscores the responsibility of company directors for fraud and misrepresentations, reaffirming that individuals who sign documents cannot disclaim personal accountability, irrespective of any preparation process involved. As a result, the decision implies that directors cannot plead the defense that any errors in AI-generated financial information stemmed from technology’s capabilities.
- Sahara India Real Estate Corp. Ltd. v. SEBI, (2012) 10 SCC 603
The Court highlighted SEBI’s far-reaching scope of powers to protect investors and ensure market integrity, justifying the broad scope of interpretation for meeting new challenges. Accordingly, the SEBI Consultation Paper on Artificial Intelligence & Machine Learning, released even without a law specifically empowering SEBI to regulate AI technology, is supported by this judicial precedent.
- Vodafone International Holdings B.V. v. Union of India, (2012) 6 SCC 613
Even though this is a case of tax law, the pronouncement by the Supreme Court regarding the “look behind” principle, which involves an inquiry into the substance rather than the form of corporate structure, can provide an analogy for the court in understanding whether the use of AI by the corporation was a substantial one or not.
CONCLUSION:
The incorporation of AI in corporate governance in India is no longer hypothetical; rather, it is functional within various boards’ reports, risk departments, and regulatory systems used by all publicly traded firms. However, the law in India still considers questions of accountability in ways suitable for human thought processes, thus widening the divergence between practice and theory.
Regulatory actions taken by SEBI in the year 2025, such as Regulation 16C and its AI/ML Consultation Paper, are commendable but only relevant to a specific sector since other sectors like banks, insurers, and general corporate governance under the Companies Act lag behind.
A way out needs three steps. First, the Ministry of Corporate Affairs should define through rules or guidelines provided under Section 166 the standards of care expected from directors when using AI recommendations.
Secondly, mandatory algorithmic audit trails and explainability requirements must apply not only to SEBI-regulated firms but to all firms that use AI in any governance function.
Thirdly, harmony between the DPDP Act, AI-specific guidance, and the company law is crucial to ensure that the fragmentation in regulation does not become a backdoor way to dodge accountability.
Finally, accountability is critical to ensuring the legitimacy of corporate governance in India. A corporate governance system that allows the delegation of decision-making powers to black-box algorithms while lacking ways to trace, audit, and attribute responsibility for such decisions threatens the fiduciary principle underpinning Indian company law.
FAQs:
Q1. Can a company director be liable for a decision reached by use of an AI algorithm?
Yes. According to Indian legislation, the liability lies with the human agents that either use or create such algorithms, thus the director cannot avoid liability based on Section 166 due to AI.
Q2. Is there a special legislation that governs AI usage in corporate governance in India?
There is no separate regulation. Currently, AI in corporate governance is governed via sector-specific regulations, including SEBI’s Regulation 16C and the 2025 AI/ML consultation paper, along with Companies Act, 2013 fiduciary principles.
Q3. What is Regulation 16C according to SEBI?
Incorporated in the SEBI (Intermediaries) Regulations, the February 2025 amendment stipulates sole liability for all AI/ML products regardless of who creates and uses them.
Q4. What is the connection between the DPDP Act of 2023 and AI governance?
According to DPDP, any processing of personal data by AI for governance-related analysis should be conducted by “data fiduciaries” that would make sure of its lawful processing, subject to penalties of up to two hundred and fifty crore rupees.
