Abstract
The Indian Law is facing a complex challenge of rise in deepfake technology. It affects constitutional rights, the reliability of evidence, and the responsibility of intermediaries. This paper looks at deepfakes, which are lifelike synthetic media created by Generative Adversarial Networks (GANs). These deepfakes violate Article 21, which protects life, liberty, dignity, and informational privacy according to Puttaswamy judgement. Unlike traditional harms, deepfakes allow non-consensual biometric duplication, undermining identity control and causing psychological harm, especially to women through sexualized or defamatory content.The paper critiques the Bharatiya Sakshya Adhiniyam, 2023 (BSA) for not properly validating AI-generated evidence beyond basic certification under Sections 62-63. This leaves courts exposed to fake manipulations, even with the limits of forensic detection. The safe harbors for intermediaries under Section 79 of the IT Act, 2000, fail because of issues with traceability and neutrality. Platforms cannot check content neutrality without making editorial decisions.The paper suggests a comprehensive framework: (1) courts should recognize digital identity rights through the horizontal application of Article 21; (2) amend the BSA for better authentication hearings and verification; (3) reform the IT Act to provide safe harbors for good-faith removals and to shift the burden of proof to the complainant; (4) specify deepfake forgery in BNS and integrate it with the DPDPA. This all-encompassing approach tackles related issues and calls for legal and judicial action to protect constitutional values in the age of AI.
Introduction
One of the biggest issues facing Indian jurisprudence today is the rise of deepfake technology, which has led to a crisis in both the constitution and the evidentiary crises that surpasses conventional legal classifications. In addition to violating Article 21 constitutional protections, deepfakes; hyper realistic synthetic media produced by Generative Adversarial Networks (GANs) and sophisticated machine learning algorithms, also undermine the authentication processes of the Indian Evidence Act and reveal significant weaknesses in intermediary liability frameworks[1]. In order to address the three-way harm caused by deepfake technology, this paper makes the case for an integrated legal framework that combines forensic authentication standards, constitutional protections, and reformed intermediary accountability.
It is impossible to estimate how serious this challenge is. Unlike traditional defamation or privacy violations, deepfakes operate at the nexus of three distinct legal domains: they attack constitutional identity rights of individuals who are non-consensually synthesized without remedial frameworks for private-actor harm; they undermine the presumption of electronic record reliability upon which Indian courts depend for evidence authentication; and they render the “neutrality” principle of intermediary liability theoretically impossible when platforms must detect synthetic media they cannot technically verify.
The present state and structure of the law ecosystem made up of fragmented provisions under the Information technology act, 2000, the Bhartiya Nyaya Sanhita, 2023, and the Indian Evidence Act, 1872 do not fully tackle this holistic crisis.
This paper takes the form of four composite analyses the first to investigate the constitutional infringement of Article 21 in terms of identity rights and informational privacy, the second one to examine the authenticity crisis of evidentiary authentication posed by deepfakes under Section 62 -63 of the Bhartiya Sakshya Adhiniyam, the third one to consider the structural inability of intermediary neutrality under Section 79 of the IT Act and lastly the fourth to suggest an all-inclusive model that integrates constitutional protection mechanisms, the standard of evidentiary authentication.
[1] Vishnu Vardhan G and Mahizhnan C, ‘Deepfakes and Digital Evidence: A New Test for Indian Courts’ (2026) 6(1) International Journal of Advanced Legal Research