Summary: Artificial Intelligence is transforming legal research, drafting and access to basic legal information, but the same technology is also enabling deepfakes, identity misuse, fraud and unauthorised exploitation of personality rights. Recent proceedings before the Delhi High Court involving celebrities and public figures illustrate how AI-generated faces, voices and likenesses can be commercially exploited or used for deceptive content, requiring courts to consider increasingly flexible remedies. Beyond celebrities, deepfake technology presents wider risks of impersonation and financial fraud. India’s evolving response includes intermediary regulation dealing with synthetically generated information, alongside existing remedies under the Bharatiya Nyaya Sanhita, 2023, the Protection of Children from Sexual Offences Act, 2012 and the Digital Personal Data Protection Act, 2023. The challenge for lawyers and citizens is therefore twofold: learning to use AI productively while also recognising manipulated digital identities and understanding the legal remedies available against their misuse.
Introduction: Picture a courtroom. Not the marble-and-mahogany kind from old movies, but a modern one, where a judge is staring at a video that looks exactly like a famous actor, except the actor never said those words, never wore that outfit, and never consented to any of it. This is not science fiction anymore. This is Tuesday in the Delhi High Court.
Artificial Intelligence has quietly slipped into every corner of our lives, including the very serious, very slow-moving world of Indian law. And like any powerful guest, it has brought gifts in one hand and a small crowbar in the other.
The Good News First: A Fresher’s Best Friend
For a law student stepping into internships and first jobs today, AI is less a threat and more a very fast, very tireless junior colleague. Case research that once meant hours flipping through bulky reporters can now be narrowed down in minutes. Drafting a first cut of a legal notice, structuring a memo, or summarising a 40-page judgement, AI tools handle the grunt work, freeing up young lawyers to focus on the part machines still cannot do: judgment, strategy, and the human reading of a client’s real problem.
The World Economic Forum’s own 2026 outlook on AI at work makes an interesting point, the disruption isn’t limited to entry-level jobs, as many feared. It’s actually mid-level roles that are feeling the heavier pressure, because a lot of “experienced but repetitive” work is exactly what AI is best at automating. For freshers, that is oddly good news: the profession will always need people willing to learn the new tools from day one, rather than unlearn old habits later.
For the common person too, AI has been democratising. Legal information that once needed an expensive consultation is now a chatbot query away: understanding a rental agreement clause, knowing your rights during a police stop, or getting a first sense of whether that termination notice from work was even lawful. Small businesses draft their first contracts with AI assistance instead of skipping legal help altogether because a lawyer felt unaffordable. That is real access to justice, even if an imperfect version of it.
Now, the Crowbar
But the same technology that helps a fresher draft a bail application can also help a stranger destroy someone’s identity in an afternoon.
R. Madhavan Deepfake Case
Take the case that shook Indian celebrity law in December 2025 — Ranganathan Madhavan v. G Filmz Studioz & Ors. Actor R. Madhavan found dozens of YouTube channels running AI-generated deepfake trailers and obscene content built entirely on his face and voice, alongside merchandise sold in his name without a rupee of consent. The Delhi High Court had to step in and order takedowns across a staggering 21 defendants, a list that read like a directory of the deepfake economy itself.
Sadhguru and Dynamic Injunctions
It wasn’t a one-off. Months earlier, in Sadhguru Jagadish Vasudev & Anr. v. Igor Isakov & Ors., the same court had to invent something almost unheard of, a “dynamic” injunction, because ordinary court orders couldn’t keep up with rogue websites cloning the spiritual leader’s voice and image to peddle fake investment schemes and bogus health advice. The judge recognised what many of us are only beginning to feel in our bones: a face and a voice are no longer proof of anything.
Deepfakes as a Wider Fraud Risk
This isn’t just a celebrity problem. Globally, engineering firm Arup lost about Dollar 25 million in 2024 after an employee was fooled on a video call by deepfaked colleagues, including a fake CFO, all convincing enough to authorise real wire transfers. Multiply that by ordinary families being scammed by a fake “kidnapped child” voice call, or a fake “boss” asking for an urgent bank transfer, and the crowbar starts to look less like an occasional inconvenience and more like a genuine crime wave.
Can the Law Keep Up?
India has tried to respond, and rather seriously. The IT (Intermediary Guidelines and Digital Media Ethics Code) Amendment Rules, 2026, notified in February, created a whole new legal category called “synthetically generated information.” The headline change: platforms must now take down unlawful deepfakes within just three hours of a court or government order, down from a lazy 36 hours earlier, and an even tighter two-hour window applies to non-consensual sexual deepfakes and child sexual abuse material. AI content must also carry a visible label and permanent, unremovable metadata showing it was machine-made.
Existing Criminal, Child Protection and Data Protection Laws
These rules don’t stand alone. Prosecutors increasingly lean on the Bharatiya Nyaya Sanhita, 2023 for cheating, forgery, identity theft and defamation charges; the POCSO Act, 2012 where children are victims; and the Digital Personal Data Protection Act, 2023 to argue that a person’s face and biometric likeness deserve the same protection as their bank details.
So, Friend or Foe?
Honestly, both, depending on who is holding the tool. For a diligent fresher or an honest small business owner, AI is a genuine equaliser, opening doors that used to need deep pockets. For a scammer or a rogue website, it is a weapon that outruns old thirty-six-hour laws in a matter of minutes.
The real test for India’s legal profession, freshers included, won’t be learning to use AI. Everyone will learn that fast, out of necessity. The real test will be learning to spot when a “person” on the other side of a screen isn’t a person at all and knowing, chapter and section, exactly what the law now lets you do about it.
The views expressed are personal.
Authors:
Abhisikta Nandy, Student, B. A. LL. B IPR Hons., KIIT School of Law, mail: abhisiktanandy295@gmail.
Daya Shankar Padhi, B. A. LL. B IPR Hons., KIIT School of Law, mail: [email protected]
 |
 |
|---|---|
| Abhisikta Nandy | Daya Shankar Padhi |
