India is drawing a line in the sand against the “wild west” of Artificial Intelligence.
Under the newly notified Information Technology Amendment Rules, 2026, the Ministry of Electronics and Information Technology (MeitY) has introduced a formal legal definition for Synthetically Generated Information.
Starting February 20, 2026, any AI-generated content that looks or sounds real enough to fool the human eye must carry a digital “ID card” or face immediate removal.
What Counts as “Synthetically Generated”?
The law defines this as any audio, visual, or audio-visual information—be it a voice note, a photo, or a video—that has been created or altered using computer resources in a way that it appears authentic.
The “litmus test” is simple: if the content portrays a person or event so realistically that it is indistinguishable from a real-world person or event, it falls under these strict new mandates.
The ‘Safe Zone’: What Isn’t a Deepfake?
Creators don’t need to panic about basic filters or editing. The government has carved out specific exemptions for routine digital work:
- Standard Edits: Color correction, noise reduction, and cropping are safe, provided they don’t change the context or meaning of the original media.
- Professional Tools: Creating PDFs, PowerPoint presentations, or using AI for language translation and accessibility is exempt.
- Quality Boosters: Tools used solely to improve clarity or searchability without manipulating the “material part” of the information are permitted.
New Mandates for Platforms and Creators
If your content is synthetic, the transparency requirements are now non-negotiable. Platforms must ensure that AI content is prominently labeled with a visible tag or an audio disclaimer. Furthermore, this content must have permanent metadata—a hidden digital watermark—tracking it back to the specific computer resource used to create it.
Significant social media platforms are now legally required to ask users for a declaration before they hit “upload”. If a platform knowingly allows a deepfake to go viral without these labels, they lose their “safe harbor” protection and become liable for the content.
Why the Crackdown?
The primary goal is to stop the weaponization of AI. The rules specifically ban the generation of non-consensual intimate imagery, child sexual abuse material, and AI-impersonation meant to deceive the public. With the response time for removing illegal content slashed from 36 hours to just 3 hours, the era of “wait and see” for deepfakes in India is officially over.