The Avatar Liability: Legal Status of AI Influencers in India

Introduction

Naina has no bank account, no address, and no birth certificate — yet she quotes six to eight lakh rupees per Instagram reel, has fronted campaigns for Oppo, and outearns most human influencers. She, like Kyra and Myntra’s Maya, is a computer-generated persona marketed as a person. When her endorsement misleads, exaggerates, or harms, who exactly is liable?

Since Kyra’s 2022 debut, India has built a real industry around AI influencers, and regulators are catching up. The Advertising Standards Council of India (ASCI) first defined “virtual influencer” in its Guidelines for Influencer Advertising in Digital Media (Aug. 2023).[1] Its 2026 revision now mandates a standalone “AI-generated” disclosure, with the deploying brand bearing primary compliance responsibility. The Consumer Protection Act, 2019, read with the Central Consumer Protection Authority (CCPA) Guidelines, 2022, treats influencers as “endorsers” facing penalties up to ten lakh rupees.[2] This regime was built for a human who can be summoned and questioned, not a persona with no body or mind to interrogate.

This article asks a narrower question: when an AI influencer’s conduct — a false claim, a fabricated endorsement, an appropriated likeness — causes harm, does Indian law fix liability on an identifiable party? It argues liability does not vanish because the endorser is fictional; it shifts, often unpredictably, onto the brand, the creator-studio, or the platform, leaving consumers and the humans behind these avatars exposed without settled rules.

The analysis is confined to civil and regulatory liability under consumer protection law, advertising self-regulation, and Information Technology Act, 2000 intermediary rules,[3] along with copyright and personality-rights questions where an avatar echoes a real individual. It excludes deep fake criminal law and comparative EU/US frameworks except where directly instructive. Part II traces the rise of AI influencers in India; Part III tests “endorser” liability under the CCPA Guidelines; Part IV examines intermediary liability, copyright authorship, and personality rights; Part V proposes where the accountability gap should close, and by whom.

Mapping the Statutory Landscape

India has no statute that uses the words “AI influencer” or “virtual persona.” What exists is a patchwork of laws never built for synthetic media, now being stretched to cover it.

Start with the Information Technology Act, 2000. Section 79 gives intermediaries safe harbour if they exercise due diligence[4], the provision platforms hosting AI-generated content rely on most. That safe harbour was tightened by the Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Amendment Rules, 2025, notified by MeitY on 22 October 2025 and effective from 15 November 2025.[5] For the first time, Indian law defines “synthetically generated information” under the newly inserted Rule 2(1)(wa). Under Rule 3(3), any intermediary enabling the creation of such content must label it permanently, covering at least ten percent of the visual frame, or the opening ten percent of an audio clip’s duration. An AI influencer’s every post now falls within this mandate, at least on paper.

Consumer protection law runs a parallel track. The Consumer Protection Act, 2019, the CCPA’s 2022 Guidelines for Prevention of Misleading Advertisements and Endorsements, and the January 2023 “Endorsement Know-Hows” require an “endorser” to disclose material connections, and specifically require virtual influencers to disclose that consumers are not interacting with a real person.[6] ASCI’s Code, amended in 2023 and tightened again in 2026, goes further: where the endorser is a virtual creation, the brand deploying it is treated as the advertiser and bears primary responsibility, since the avatar plainly cannot be made to answer for its own claims.[7]

The Digital Personal Data Protection Act, 2023, whose rules were notified only on 13 November 2025 with compliance staggered through May 2027, becomes relevant the moment an avatar’s face or voice is trained on a real individual’s data without consent, more often than most brands admit.[8]

What is missing entirely is any statute recognising personality or publicity rights. India has built this protection purely through Article 21 privacy jurisprudence and common law, which is where the courts come in.

III. Judicial Interpretation: How Courts Have Filled the Vacuum

Three decisions between 2023 and 2024 now anchor every discussion of AI and identity in India, and it is worth being precise about what each actually held, because their reach is narrower than most commentary suggests.

In Anil Kapoor v. Simply Life India & Ors. (Delhi High Court, 2023),[9] Justice Pratibha M. Singh granted a sweeping injunction protecting the actor’s name, voice, image, and catchphrase “Jhakaas” against AI deepfakes, morphed GIFs, and unauthorised merchandise, grounding personality rights in the Article 21 dignity guarantee and treating persona as commercially valuable property.

Jaikishan Kakubhai Saraf alias Jackie Shroff v. The Peppy Store & Ors. (Delhi High Court, May 2024)[10] went further. Justice Sanjeev Narula restrained not just merchandise and morphed videos but an unlicensed AI chatbot built to mimic Shroff’s voice and his signature “Bhidu,” the first time an Indian court confronted a conversational AI system trained on a celebrity’s persona.

Arijit Singh v. Codible Ventures LLP (Bombay High Court, July 2024)[11] pushed the doctrine further still. Justice R.I. Chagla restrained AI platforms from synthesising the singer’s voice, condemning the practice in unusually strong terms, and extended protection to his vocal style, mannerisms, and signature, attributes no earlier judgment had protected with this precision.

Read together, these cases confirm one thing clearly: Indian courts will protect a real person against an AI system built in their image.[12] What none of them decide is the question this article is actually concerned with, what happens when the avatar is not modelled on any identifiable individual, and it is the avatar’s own conduct, a false claim, a defamatory remark, a fraudulent endorsement, that causes harm to a third party. On that question, Indian jurisprudence has nothing to say yet.

The Liability Gap: A Critical Evaluation

This is where description has to give way to argument. An AI influencer, however lifelike, has no legal personality. It cannot be sued, cannot hold assets, and cannot be made to comply with an injunction, a point Indian commentary on endorsement liability has already conceded. Every obligation that appears to attach to the avatar must, as a matter of legal necessity, attach instead to whichever human or corporate principal stands behind it.

The trouble is that Indian law has not worked out who that principal is. Is it the studio that designed the character? The brand that scripts its captions and signs its endorsement deals? The platform hosting it? The developer who trained the underlying model? The CCPA’s endorsement framework answers this only partially, by making the deploying brand primarily responsible, and only within the narrow context of paid endorsements.[13] It is silent on defamation committed by an AI persona, silent on fraudulent financial advice dispensed through a virtual “Influencer,” and unhelpful on who owns what the avatar creates. The Copyright Act, 1957 attributes authorship of a computer-generated work to “the person who causes the work to be created” under Section 2(d)(vi),[14] a formula drafted in 1994 for a single programmer at a keyboard, not for the layered chain of model developer, platform, and brand standing behind a typical AI influencer. Ownership, like liability, ends up nowhere in particular.

The IT Rules’ 2025 labelling mandate[15] solves a transparency problem, not a liability one. Telling a consumer that content is synthetically generated does not tell them whom to sue when that content misleads or defames them. Until Indian law expressly names the “deploying entity” as the party answerable for an avatar’s conduct, liability will keep being litigated the hard way, through personality-rights suits that work only when a real person’s likeness has been stolen, and never through a framework built specifically for synthetic personas that answer to no one but their creators.

Comparative Perspectives

Other jurisdictions have moved past this ambiguity. Article 50 of the European Union’s AI Act, directly enforceable since 2 August 2026, requires that AI-generated audio, image, video, or text be marked as such, and specifically obliges “deployers,” not merely developers, to disclose when a person is interacting with a chatbot or with synthetic content.[16] Liability is anchored to whoever actually puts the system to use.

Tennessee’s ELVIS Act, in force since July 2024, offers a sharper model still.[17] It extended the state’s existing right-of-publicity statute to voice and created secondary civil liability for anyone who distributes an AI-generated voice or likeness knowing it was unauthorised, reaching past the original creator to the platforms and distributors that profit from the content. China’s deep-synthesis regulations move in a similar direction, requiring registration and labelling from providers rather than end users.[18] India’s own ASCI framework gestures toward the same idea by fixing responsibility on the deploying brand,[19] but it remains a self-regulatory code, stiffened only by CCPA’s enforcement muscle rather than backed by a statute of general application.

The lesson is not that India should copy either model, but that both have done what Indian law has not: named, in the text of a binding rule, exactly who answers for the avatar.

Conclusion — Locating the Liable Party

The legal status of AI influencers in India can be stated plainly: they have none. No statute grants them personality, and none imposes obligations on them directly, because an algorithm cannot be summoned, sued, or made to pay a fine. Everything this article has traced, the labelling mandate under the 2025 IT Rules, the endorsement disclosures under the CCPA and ASCI framework, the personality-rights injunctions in Anil Kapoor, Jackie Shroff, and Arijit Singh[20], in fact regulates the humans and companies standing behind the avatar, not the avatar itself. That is the article’s central finding, and it answers the question this piece set out to ask.

India’s law does not need to decide whether an AI influencer can be liable; it needs to decide, clearly and in advance, which human or corporate principal answers for it. At present that answer exists only in fragments: the CCPA names the brand, but only for paid endorsements; the courts name the actual celebrity being impersonated, but only when a real person’s face or voice has been stolen. Neither reaches the more common scenario, a wholly fictional persona that misleads, defames, or defrauds on its own account.

The practical stakes are immediate. Brands are already deploying virtual influencers in India’s creator economy without a settled compliance checklist, and consumers interacting with them have no obvious party to hold responsible when something goes wrong. Closing that gap does not require inventing new legal categories. It requires India to do what other jurisdictions discussed above have already done: name the “deploying entity” as the responsible party across contexts, not just endorsements, and extend personality and publicity protection beyond the reactive, case-by-case route this article has documented.

The avatar will keep getting more convincing. The law’s task was never to regulate the illusion, only to make sure a human hand is always find-able behind it.

Table of Authorities

Cases

Anil Kapoor v. Simply Life India & Ors., 2023 SCC OnLine Del 6914 (Del. H.C. Sept. 20, 2023).

Arijit Singh v. Codible Ventures LLP, 2024 SCC OnLine Bom 2445 (Bom. H.C. July 26, 2024).

Jaikishan Kakubhai Saraf alias Jackie Shroff v. The Peppy Store & Ors., 2024 SCC OnLine Del 3664 (Del. H.C. May 15, 2024).

Constitutional Provisions

India Const. art. 21.

Statutes and Regulations — India

Consumer Protection Act, No. 35 of 2019, India Code (2019).

Copyright Act, No. 14 of 1957, India Code (1957), § 2(d)(vi).

Digital Personal Data Protection Act, No. 22 of 2023, India Code (2023).

Information Technology Act, No. 21 of 2000, India Code (2000), § 79.

Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Amendment Rules, 2025, Gazette of India, pt. II, § 3(i) (Oct. 22, 2025).

Central Consumer Protection Authority, Guidelines for Prevention of Misleading Advertisements and Endorsements, 2022, Gazette of India, pt. III, § 4 (June 9, 2022).

Department of Consumer Affairs, Endorsement Know-Hows (Jan. 2023).

Self-Regulatory Codes

Advertising Standards Council of India, Guidelines for Influencer Advertising in Digital Media (Aug. 2023, as amended 2026).

Advertising Standards Council of India, Code for Self-Regulation in Advertising (2023, revised 2026).

Foreign and International Materials

Regulation (EU) 2024/1689 of the European Parliament and of the Council of 13 June 2024 Laying Down Harmonised Rules on Artificial Intelligence, art. 50, 2024 O.J. (L 1689).

Tenn. Code Ann. §§ 47-25-1101 to -1112 (2024) (Ensuring Likeness, Voice, and Image Security Act).

Cyberspace Administration of China, Provisions on the Administration of Deep Synthesis Internet Information Services (2022, amended 2023).

[1]Advertising Standards Council of India, Guidelines for Influencer Advertising in Digital Media (Aug. 2023), https://ascionline.in.

[2]Consumer Protection Act, No. 35 of 2019, India Code (2019); Central Consumer Protection Authority, Guidelines for Prevention of Misleading Advertisements and Endorsements, 2022, Gazette of India, pt. III, § 4 (June 9, 2022).

[3]Information Technology Act, No. 21 of 2000, India Code (2000), § 79.

 

 

[5]Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Amendment Rules, 2025, Gazette of India, pt. II, § 3(i) (Oct. 22, 2025), r. 2(1)(wa), r. 3(3).

[6]Consumer Protection Act, No. 35 of 2019, India Code (2019); Central Consumer Protection Authority, Guidelines for Prevention of Misleading Advertisements and Endorsements, 2022, supra note 2; Department of Consumer Affairs, Endorsement Know-Hows (Jan. 2023), https://consumeraffairs.nic.in.

[7]Advertising Standards Council of India, Code for Self-Regulation in Advertising (2023, revised 2026), https://ascionline.in.

[8]Digital Personal Data Protection Act, No. 22 of 2023, India Code (2023) (rules notified Nov. 13, 2025, Gazette of India, pt. II, § 3(i)).

[9]Anil Kapoor v. Simply Life India & Ors., 2023 SCC OnLine Del 6914 (Del. H.C. Sept. 20, 2023).

[10]Jaikishan Kakubhai Saraf alias Jackie Shroff v. The Peppy Store & Ors., 2024 SCC OnLine Del 3664 (Del. H.C. May 15, 2024).

[11]Arijit Singh v. Codible Ventures LLP, 2024 SCC OnLine Bom 2445 (Bom. H.C. July 26, 2024).

[12]Anil Kapoor, supra note 8; Jackie Shroff, supra note 9; Arijit Singh, supra note 10.

[13]Central Consumer Protection Authority, Guidelines for Prevention of Misleading Advertisements and Endorsements, 2022, supra note 2.

[14]Copyright Act, No. 14 of 1957, India Code (1957), § 2(d)(vi).

[15]Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Amendment Rules, 2025, r. 3(3), supra note 4.

[16]Regulation (EU) 2024/1689 of the European Parliament and of the Council of 13 June 2024 Laying Down Harmonised Rules on Artificial Intelligence, art. 50, 2024 O.J. (L 1689).

[17]Tenn. Code Ann. §§ 47-25-1101 to -1112 (2024) (Ensuring Likeness, Voice, and Image Security Act).

[18]Cyberspace Administration of China, Provisions on the Administration of Deep Synthesis Internet Information Services (2022, amended 2023), http://www.cac.gov.cn.

[19]Advertising Standards Council of India, Code for Self-Regulation in Advertising, supra note 6.

[20]Anil Kapoor, supra note 8; Jackie Shroff, supra note 9; Arijit Singh, supra note 10.