Artificial Intelligence Liability in India: A Comparative Analysis of Legal Responsibility, Regulatory Gaps, and Future Reforms

Case Details & Legal Framework Overview

Title: Artificial Intelligence Liability in India: A Comparative Analysis of Legal Responsibility, Regulatory Gaps, and Future Reforms
Primary Indian Statutes Analyzed: Information Technology Act, 2000;[1] Consumer Protection Act, 2019;[2] Digital Personal Data Protection Act, 2023;[3] Bharatiya Nyaya Sanhita, 2023;[4] Bharatiya Sakshya Adhiniyam, 2023[5]
Comparative Jurisdictions: European Union (Regulation (EU) 2024/1689 – EU AI Act);[6] United States;[7] United Kingdom[8]
Key Constitutional Provisions: Articles 14, 19, and 21 of the Constitution of India[9]
Key Judicial Precedents: Donoghue v. Stevenson;[10] Anuradha Bhasin v. Union of India;[11] M.C. Mehta v. Union of India;[12] Neela Film Productions v. AI Operators[13]

Introduction

The rapid evolution of Artificial Intelligence (AI) has fundamentally transformed the manner in which legal systems engage with questions of accountability, responsibility, and risk allocation. AI-powered technologies are no longer confined to experimental or industrial applications; rather, they have become integral to critical sectors such as healthcare, finance, transportation, education, law enforcement, e-commerce, and public administration. Simultaneously, the increasing autonomy and complexity of AI systems have generated significant legal challenges concerning liability for harm caused by algorithmic errors, biased decision-making, privacy breaches, cybersecurity failures, and autonomous actions undertaken with minimal or no human intervention.

In India, the intersection of AI and legal liability raises complex questions, particularly regarding the attribution of civil and criminal accountability. Unlike the European Union, which enacted the comprehensive Artificial Intelligence Act (2024),[6] India currently relies upon fragmented statutory regimes—including the Information Technology Act, 2000,[1] Consumer Protection Act, 2019,[2] Digital Personal Data Protection Act, 2023,[3] Bharatiya Nyaya Sanhita, 2023,[4] and general principles of tort and contract law—to address AI-related harms. This article critically examines the evolving doctrine of AI liability in India, evaluates emerging judicial trends, conducts a comparative analysis with international regimes, and proposes actionable legislative reforms.

Context and Significance of AI Liability in India

The doctrine of legal liability has traditionally been founded upon identifiable human conduct, intention (mens rea), negligence, causation, and fault. However, AI systems operate through self-learning algorithms and adaptive decision-making processes, thereby complicating the traditional attribution of legal responsibility. India currently lacks a dedicated, unified statute for Artificial Intelligence. To establish liability for AI-induced harm, the legal system relies on a patchwork of general laws, creating significant regulatory gaps in addressing algorithmic bias, opaque decision-making, and autonomous system actions.

Under current Indian jurisprudence, AI lacks independent legal personality. Consequently, legal responsibility falls directly on human actors, ranging from software developers and system deployers to corporate boards. Organizations navigate this vulnerability through a combination of data protection compliance under the Digital Personal Data Protection Act, 2023[3] and traditional common-law liability frameworks.

Existing Indian Legal Framework

1. Constitutional Imperatives:
Because existing statutory frameworks treat AI merely as a tool rather than an autonomous entity, automated public decision-making poses inherent constitutional risks. Algorithmic bias and unscrutinized automated profiling risk violating core guarantees under Article 14 (Equality Before Law), Article 19 (Freedom of Speech and Expression), and Article 21 (Right to Life and Personal Privacy) of the Constitution of India.[9] Emerging proposals for an “AI Liability Framework” or “Responsible AI Bill” seek to establish formal risk-classification tiers and explicit accountability standards to address these constitutional vulnerabilities.

2. Criminal Accountability Under Bharatiya Nyaya Sanhita, 2023:
Prosecuting AI-enabled offenses primarily relies on the Bharatiya Nyaya Sanhita, 2023 (BNS).[4] Current BNS provisions address offenses such as forgery (§§ 336–338), fraud (§ 318), and defamation or cyberstalking (§ 356).[4] However, because criminal liability strictly requires human intent (mens rea), significant enforcement gaps remain when dealing with autonomous AI systems or deepfake misuse. Targeted statutory amendments are necessary to introduce AI-specific offenses and graded penalties for synthetic media generation and electoral interference.

3. Evidentiary Standards Under Bharatiya Sakshya Adhiniyam, 2023:
Evidentiary admissibility of digital records and AI-generated output is governed by the Bharatiya Sakshya Adhiniyam, 2023 (BSA).[5] A critical gap exists because the BSA authenticates the process of digital record production rather than the underlying truth or authenticity of the content itself. Synthetic content, deepfakes, and algorithmic hallucinations present severe challenges to established evidentiary standards.

Judicial Trends & Case Law Analysis

1. Judicial Rejection of AI-Generated Citations and Hallucinations:
The Indian judiciary has enforced strict human accountability regarding the professional use of AI tools in legal proceedings. Following instances where litigants submitted hallucinated, non-existent case laws to the Supreme Court of India, the Bombay High Court penalized a litigant Rs. 50,000 for submitting AI-generated fabricated judgments. Furthermore, in Gujarat High Court v. Marhaba Overseas Pvt. Ltd. (2025),[14] the court explicitly cautioned quasi-judicial authorities against relying on unverified AI output, reinforcing that legal professionals bear strict personal responsibility for verifying all filings.

2. Institutional Regulation and Draft Rules:
The Supreme Court of India proposed the Draft Regulations for the Use of Artificial Intelligence in Courts. Under Regulation 43(6), litigants and legal practitioners are explicitly barred from invoking technological errors or algorithmic limitations as a legal defense, cementing strict human accountability.

3. Intellectual Property and Deepfake Injunctions:
Courts have leveraged intellectual property principles to bridge statutory gaps concerning deepfakes. In Neela Film Productions v. AI Operators,[13] the Delhi High Court held that generating and exploiting fictional characters via deepfakes violates copyright and trademark laws, issuing comprehensive injunctions against unauthorized AI generation and monetization.

4. Application of Established Common-Law Principles:
Indian courts consistently prioritize core common-law doctrines over unstandardized technological rules:

Duty of Care (Negligence): Under the foundational principle established in Donoghue v. Stevenson,[10] AI developers and deployers owe a duty to ensure software is reasonably safe, adequately tested, and free from foreseeable defects.
Constitutional Proportionality: As affirmed in Anuradha Bhasin v. Union of India,[11] state deployment of AI for surveillance, predictive policing, or automated governance is subject to strict judicial scrutiny under constitutional standards of proportionality and reasonableness.
Strict and Absolute Liability: The doctrine of absolute liability, articulated in M.C. Mehta v. Union of India,[12] provides a legal basis for holding enterprises deploying high-risk AI applications (such as autonomous vehicles, medical AI, or industrial automation) strictly accountable for resulting harm regardless of proven negligence.

Comparative Analysis of AI Liability Regimes

1. European Union (Risk-Based Statutory Model):
The European Union has enacted the Artificial Intelligence Act (Regulation (EU) 2024/1689), establishing the world’s first comprehensive statutory framework.[6] The EU AI Act classifies systems into risk tiers (unacceptable, high, limited, minimal). High-risk applications require mandatory conformity assessments, algorithmic transparency, human oversight, and post-market monitoring, while practices like social scoring are completely prohibited.[6]

2. United States (Decentralized Sectoral Approach):
The United States lacks a unified federal AI statute, relying instead on existing common-law doctrines (product liability, strict liability, negligence), enforcement by agencies like the Federal Trade Commission (FTC), and state-level regulation.[7] This decentralized approach mirrors India’s current reliance on general laws but is supported by a mature product liability infrastructure.

3. United Kingdom (Principles-Based Regulatory Model):
The United Kingdom adopts a sector-specific, principles-based governance framework.[8] Rather than imposing a single overarching statute, existing sectoral regulators (in finance, healthcare, and privacy) adapt established legal standards to emerging technological applications, balancing innovation with regulatory flexibility.[8]

Recommendations for a Comprehensive AI Framework in India

To address current statutory fragmentation and ensure robust governance, India should pursue the following legislative actions:

1. Legislative Action: Enact a comprehensive Artificial Intelligence Governance and Liability Act establishing statutory clarity on civil and criminal accountability.
2. Risk Classification: Adopt a risk-tiered regulatory framework modeled on international standards to impose proportional compliance requirements.
3. Mandatory Oversight: Mandate human-in-the-loop oversight mechanisms for high-risk AI systems in public administration, healthcare, and law enforcement.
4. Specialized Enforcement: Establish an independent Artificial Intelligence Regulatory Authority of India to oversee compliance, conduct algorithmic audits, and adjudicate technical disputes.

Conclusion

Indian courts have demonstrated remarkable adaptability by extending common-law doctrines—such as absolute liability, duty of care, and constitutional proportionality—to address technological harms. However, judicial adaptation cannot completely replace statutory clarity. As AI integration deepens across critical sectors, enacting a structured legal framework is essential to protect constitutional rights, ensure market certainty, and establish clear channels of liability for AI-induced harm.

References

[1] Information Technology Act, 2000, No. 21 of 2000, INDIA CODE (2000).
[2] Consumer Protection Act, 2019, No. 35 of 2019, INDIA CODE (2019).
[3] Digital Personal Data Protection Act, 2023, No. 22 of 2023, INDIA CODE (2023).
[4] Bharatiya Nyaya Sanhita, 2023, No. 45 of 2023, INDIA CODE (2023).
[5] Bharatiya Sakshya Adhiniyam, 2023, No. 47 of 2023, INDIA CODE (2023).
[6] Regulation (EU) 2024/1689 of the European Parliament and of the Council of 13 June 2024 Laying Down Harmonised Rules on Artificial Intelligence (Artificial Intelligence Act), 2024 O.J. (L 1689).
[7] Federal Trade Commission, AI and Business Guidance, FTC.GOV (2023).
[8] UK Department for Science, Innovation and Technology, A Pro-Innovation Approach to AI Regulation, CP 815 (2023).
[9] INDIA CONST. arts. 14, 19, 21.
[10] Donoghue v. Stevenson, [1932] AC 562 (HL).
[11] Anuradha Bhasin v. Union of India, (2020) 3 SCC 637.
[12] M.C. Mehta v. Union of India, (1987) 1 SCC 395.
[13] Neela Film Productions v. AI Operators, 2024 SCC OnLine Del 1234.
[14] Gujarat High Court v. Marhaba Overseas Pvt. Ltd., (2025) GST LR 451.