Sakkcham Singh Parmaar is a 3rd Year student at Jindal Global Law School, O.P. Jindal Global University.
India’s Deepfake Dilemma: Navigating the New Frontier of AI Content Regulation
As the global community wrestles with this monster called synthetic content, all countries find themselves in a conundrum trying to hack the legislation thereof. For its part, India is taking a compelling step to pioneer the setting of legal fences for an age increasingly dominated by artificial intelligence. In this advance, the Ministry of Electronics and Information Technology (MeitY), during the announcement on October 22, 2025, issued drafts and recommendations for changes in the change process in Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Rules, 2021 that govern the community guidelines that seek to minimize the harms associated with the use of AI in deepfake and synthetic media production in society. It is one of the most formal steps by which India moves toward framing a structured AI governance framework, focusing on both regulatory progress of new technologies and the urgent need for user protection and digital transparency.
Catalysts for Regulation
The Indian government did not act in a vacuum but had combined rising public concern and judicial intervention as its catalyst. The “significant growth” in generative AI tool misuse-for spreading misinformation, manipulating elections, or impersonating individuals-was among the key motivators cited by the ministry. These scandals included high-profile incidents such as viral deepfake footage of actor Rashmika Mandanna in 2023. This further prompted Prime Minister Narendra Modi himself to call them a new “crisis” and led to the question becoming one of national prominence.
The public has suffered an onslaught of pressure from the judiciary. The Delhi High Court noted in the matter of Sadhguru Jagadish Vasudev & Anr. v Igor Isakov & Ors. what specific risks artificial content poses while underscoring the need for definitively clearer regulatory frameworks as well as accountability of intermediaries. Such legal nudge, together with the growing potential of AI, converted into a “weapon” to spread misinformation, harm reputation, manipulate or influence elections, or perpetrate financial fraud, created a compelling imperative for MeitY to act.`9oo0
The Draft Rules Deconstructed
On the basis of three major cores, that is, clarity, compulsory labelling, and platform obligations, the suggested amendments lay down an exhaustive framework.
Defining the “Synthetic”
First-ever legal definition in Indian law, “synthetic generation of information” was defined in the first draft rules. It states that “artificially or algorithmically created information generated, modified, or altered by using a computer resource in such a manner that information reasonably appears as its authentic or true.”
Since it’s very broad, it also allows for the inclusion of an entire universe of AI-generated content, such as video, audio, images, and text, within the ambit of the regulatory framework.
Obligation of Clear Labelling
One of the most significant features in the new rules is the imprint and clear embossing requirement. Intermediaries are required to ensure that “every such information piece is prominently labelled or embedded with a permanent unique metadata or identifier“, while the draft requires that an intermediary provide any of the tools to create or modify synthetic content. The specification is detailed for visual content, the label must cover at least 10 % of the surface area of the display, and for audio content, it must be made audible during the first 10% of its duration.
Transferring the Burden onto Platforms
It significantly improves the due diligence obligations of social media intermediaries and “Significant Social Media Intermediaries” (SSMIs), defined as large platforms with many users. Such a platform really should discharge a dual responsibility: “allow users to self-declare if the content they upload is AI-generated” and “for the most critical, deploy reasonable and appropriate technical measures” including automated tools for the verification of those user declarations and proactive detection and labelling of content that lacks such declarations. Failure to comply with these provisions would result in the loss of safe harbour protection conferred by Section 79 of the IT Act, which shields such intermediaries from liability associated with any third-party content.
Implementation Challenges and the Dissenting Voices
While there is a general agreement that regulations in this stream are supposed to be a positive step, the implementation and the unintended consequences have evoked some concerns from legal luminaries and industry representatives.
One of the primary difficulties is the technical feasibility and enforcement of the rules. Varun Singh of Foresight Law Offices has pointed out that “implementation may remain challenging as the enforcement agencies have a very poor understanding of such deepfakes and cannot detect them at an early stage“. In the end, the requirement for reasonable and appropriate technical measures presents the whole issue in a rather ambiguous manner. The evolving nature of deepfake technology ensures that the supposed detection mechanisms often trail behind.
Also, the wide definition of synthetic content is causing anxiety regarding its overreach. Arun Prabhu, partner at Cyril Amarchand Mangaldas, cautions that “by defining synthetic content without including an element of intent to deceive, the rules could force overcorrection and have a chilling effect on genuine content that has merely been cleaned up or improved.” This may affect some benign, artistic, or satirical content created through AI-assisted editing tools.
The requirement posited for the 10% visible label has its own constructive criticism in that it may interfere with Aesthetic/Artistic uses of the areas concerned by generative AI. Some of the experts propose more adaptive watermarking to fit the bill.
India in a World Grappling with AI Governance
By adopting the regulatory options, India joins an expanding class of nations that are working out norms for synthetic media. It is proposed to make the rules in line with global norms for mandatory labelling. The EU AI Act presents a requirement for AI providers to label synthetic media in a machine-readable way for detection as artificial. China has also rolled out rules requiring content providers to clearly flag AI-generated material.
However, the approaches are not the same. Denmark, for instance, might be considering a radically different model whereby individuals are granted copyright over their own likenesses and can then seek removal of digitally-altered media created without their consent. With its transparency- and platform-liability-based model, the draft framework of India positions itself as a potentially key contributor to the global discourse on AI governance.
The Road Ahead
MeitY has called upon various stakeholders and the general public to give feedback on the draft amendments until November 6, 2025. This consultation phase will thus be very important for dealing with the challenges mentioned. As Kapil Chaudhary of Dentons Link Legal put it, “Inviting stakeholder feedback is crucial to ensuring that the regulatory framework balances technological growth with necessary safeguards for online content integrity”.
In the end, the success of the regulations depends on the development of “criteria for proper implementation and collaborative frameworks between government and industry.” If amended wisely, these could push India into the limelight as a model jurisdiction for responsible AI governance, creating a calibrated system within safer and more accountable digital ecosystems where deepfakes are flagged as transparent artefacts.