India’s landmark anti-deepfake law: New criminal penalties planned after deepfakes of three army chiefs during Operation Sindoor

India is preparing a comprehensive new anti-AI deepfake law to close critical gaps in the existing legal framework, driven by a surge in cyber fraud and deepfake videos that targeted senior military officers — including multiple Army Chiefs — with a particularly intense wave of AI-generated disinformation during the 2025 India-Pakistan conflict (Operation Sindoor) .

The planned new law and the gaps it aims to close

What is being proposed

India is preparing a standalone law to combat AI deepfakes and impersonation scams. The proposal specifically aims to "close a gap in India's laws by clearly defining punishments concerning deepfakes and other related offences" and would also cover "digital arrest" fraud — a scam where victims are conned via video calls from impersonators posing as police or government officials .

What existing law lacks

Currently, deepfake-related offences are handled under scattered provisions of the Information Technology Act, 2000 — such as Section 66C (identity theft), Section 66D (cheating by impersonation), Section 66E (privacy violations), and Sections 67/67A (obscene or sexually explicit content) . No provision explicitly defines or penalises "deepfake" or "synthetically generated content" as a distinct offence, creating enforcement and prosecution gaps.

Already enacted — the IT Amendment Rules 2026

In February 2026, the government also notified the Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Amendment Rules, 2026, which took effect on 20 February 2026 . These rules introduce the legal category of "Synthetically Generated Information" (SGI) and impose new obligations on intermediaries:

• Mandatory labelling and traceable metadata for AI-generated content
• A 3-hour takedown deadline for platforms like Facebook, Instagram, X, and Telegram to remove content flagged as harmful deepfakes
• Reasonable technical measures to prevent the hosting or dissemination of SGI that could cause harm

These rules are an operational first step, but the government acknowledges they do not replace the need for a dedicated criminal law that creates standalone deepfake offences with clear punishments .

Deepfake surge targeting top military officers

The push for tougher law has been intensified by a wave of deepfakes directly targeting India's top generals:

Three Army Chiefs deepfaked in early 2026

In the first three months of 2026 alone, three different Indian Army Chiefs were impersonated via deepfake videos . In March 2026, a viral clip showed General Upendra Dwivedi apparently admitting that India shared an Iranian warship's coordinates with Israel — which the PIB Fact Check and independent AI detection tools confirmed was AI-generated .

Deepfakes during and after Operation Sindoor (May 2025)

Following India's launch of Operation Sindoor on 7 May 2025, Pakistan launched a "coordinated information war" across social media, using AI-generated content, deepfakes, fake images, and even video game footage to spread false narratives . The Indian Army disclosed that Pakistan deployed AI tools, deepfakes, and coordinated influencer campaigns to manipulate public perception within minutes of military developments .

Continuing post-conflict attacks

As recently as July 2026, a deepfake showed Chief of Army Staff General Dhiraj Seth accusing his own institution of covering up the deaths of Operation Sindoor martyrs . Another deepfake targeted Defence Minister Rajnath Singh . PIB Fact Check units have repeatedly debunked AI-generated videos falsely attributed to Lt Gen Manoj Kumar Katiyar (Western Command Chief), Lt Gen Devendra Sharma, and other senior officers .

Key concerns deepened by the crisis

AI as a weapon of disinformation in conflict

The scale and speed of deepfake deployment during Operation Sindoor showed that AI-generated content can now be used as a real-time "weapon of mass distortion" during active hostilities, eroding trust in official military communications .

Platforms cannot keep up

Despite the 3-hour takedown rule, many deepfake videos remain online long enough to go viral, especially when coordinated state-backed networks amplify them .

Risk to national security

Deepfakes falsely showing senior generals making admissions or accusations — about troop morale, foreign operations, or war casualties — risk triggering diplomatic incidents, public panic, or operational confusion .

Inadequate deterrence

Without clear criminal penalties specifically for creating or distributing deepfakes, prosecutors must rely on stretched interpretations of existing cybercrime sections, which delays justice and weakens deterrence .