20 February 2026 se India mein AI-generated content, jise law "synthetically generated information" kehta hai, pe label lagana legally zaroori ho gaya hai. AI content post karna illegal nahi hai, lekin bina label ke post karna rules ka violation hai. Harmful deepfakes ke liye takedown window 36 ghante se ghata kar roughly 3 ghante kar diya gaya hai.
10 February 2026 ko MeitY ne Gazette notification G.S.R. 120(E) ke through IT Amendment Rules, 2026 notify kiye, aur sirf 10 din baad, 20 February se ye rules effective ho gaye. Platforms ko compliance ke liye barely koi runway mila.
Agar tum AI se thumbnails, voiceovers, videos, ya images banate ho, ye rules directly tumhe affect karte hain. Is guide mein hum simple bhasha mein samjhenge kya badla hai, creators ko practically kya karna hai, aur kaunsi cheezein ab genuinely risky hain.
**Note:** Ye general information hai, legal advice nahi. Specific compliance decisions ke liye qualified lawyer se consult karo.
Naye Rules Mein "Synthetically Generated Information" Kya Hai?
Pehli baar Indian law mein AI content ki ek clear legal definition aayi hai. SGI ka matlab hai aisa content jo artificially ya algorithmically create, generate, modify, ya alter kiya gaya ho, aur jo real, authentic, ya true lage.
Ye definition kaafi wide hai, deepfakes, AI-cloned audio, synthetic video, AI-generated images, sab isi ke andar aate hain. Matlab ek AI thumbnail se le kar AI voiceover tak, sab is framework mein cover ho sakta hai.
Important baat: ye ek naya standalone law nahi hai. Ye 2021 ke existing IT Intermediary Rules mein amendment hai, matlab pehle se maujood framework mein AI content ko add kiya gaya hai.
Kya AI Content Post Karna Ab Illegal Hai?
Nahi, ye sabse bada misunderstanding hai jo phail raha hai. AI content banana aur post karna poori tarah legal hai. Jo mandatory hua hai wo hai labelling, matlab clearly batana ki ye content AI se bana hai.
Lekin kuch content label ke saath bhi illegal rehta hai: defamatory content, kisi ki impersonation, aur non-consensual intimate images. Ye cheezein label lagane se legal nahi ho jaati, in pe alag se action ho sakta hai.
Simple rule: AI use karo, bas chhupao mat. Transparency hi poore framework ka core hai.
Label Kaisa Hona Chahiye?
Visual content pe label prominent, easily noticeable, aur adequately perceivable hona chahiye. Audio content ke liye, shuruaat mein ek clearly prefixed spoken disclosure chahiye, matlab audio ke start mein bataya jaaye ki ye synthetic hai.
Ek important clarification: draft rules mein ek proposal tha ki label screen ke 10% area ka hona chahiye, lekin ye final notified rules mein shamil nahi hua. Ab law exact size fix nahi karta, bas "prominent aur easily noticeable" chahiye.
Platforms ko provenance metadata bhi preserve karna hota hai, jo content ki origin trace karne mein help karta hai, aur ye metadata ya label remove ya obscure karna allowed nahi hai.
Takedown Ke Rules Kitne Sakht Ho Gaye Hain?
Sabse bada change speed ka hai. Court order ya authorised government notice ke through flagged unlawful content ke liye, intermediaries ko roughly 3 ghante ke andar content hatana hota hai, pehle ye window 36 ghante ka tha.
Non-consensual intimate image ki complaint ke liye ye aur bhi sakht hai, roughly 2 ghante ke andar action lena hota hai. Baaki prohibited content categories ke liye alag timelines apply hote hain.
Ye timelines mainly platforms pe apply hote hain, individual creators pe nahi. Lekin agar tumhara content is category mein aata hai, wo bahut tezi se hat sakta hai, aur appeal karne ka time bahut kam milega.
Platforms Pe Kya Obligations Aaye Hain?
Bade platforms ko ab uploaders se declare karwana hota hai ki content AI-generated hai ya nahi, aur un declarations ko verify bhi karna hota hai reasonable technical measures ke through.
Safe harbour, matlab Section 79 ka wo protection jisse platforms users ke content ke liye directly liable nahi hote, ab conditional ho gaya hai. Agar platform labelling ya takedown obligations meet nahi karta, ye protection kho sakta hai.
Yehi wajah hai ki platforms creators pe strict ho rahe hain. X ne "Made with AI" labels roll out kiye, aur baaki platforms bhi apne detection systems tight kar rahe hain, kyunki unka apna legal exposure isse linked hai.
Creators Ko Practically Kya Karna Chahiye?
Har AI-generated ya AI-modified content pe clearly label lagao, chahe platform khud auto-detect kare ya na kare. Apni taraf se transparency rakhna safest approach hai.
Platform ka AI disclosure toggle use karo agar available ho, jaise YouTube, Instagram, aur X sab ab apne upload flow mein ye option dete hain.
AI voiceover use kar rahe ho toh audio ke start mein disclosure add karo, ya video description aur on-screen text dono mein clearly mention karo.
Kisi real person ka face, voice, ya likeness bina permission ke AI se recreate mat karo. Ye labelling se cover nahi hota, ye alag se impersonation ke under aa sakta hai.
Kaunsi Cheezein Ab Genuinely Risky Hain?
Kisi celebrity ya public figure ka AI face ya voice use karke content banana, chahe parody ke naam pe, ab zyada risky hai. Impersonation aur defamation dono categories mein ye aa sakta hai.
AI-generated news-style content jo real ghatna jaisa lage, misinformation category mein problem create kar sakta hai, especially agar label na ho.
Rules mein ye provision bhi hai ki agar AI ke through koi crime hota hai, platforms ko law enforcement ke saath creator ki identity share karni pad sakti hai. Anonymity ab utni protection nahi deti jitni pehle thi.
Ye Rules Kis Pe Apply Hote Hain?
Rules un sabhi intermediaries pe apply hote hain jo Indian users ko serve karte hain, chahe company India mein registered ho ya nahi. Social media platforms, video sharing sites, messaging services, hosting providers, aur kuch AI tools bhi isme aate hain.
Individual creators direct regulated entities nahi hain is framework mein, obligations mainly platforms pe hain. Lekin practically, platforms ye obligations apne users tak pass karenge, isliye creators ko bhi comply karna hi padega.
Aage Kya Ho Sakta Hai?
Ye rules constitutional challenges ka saamna kar rahe hain, aur Digital India Act, jo ek bada legislative framework hai, abhi pipeline mein hai. Matlab ye final picture nahi hai, aur cheezein aur badal sakti hain.
Creators ke liye practical advice: compliance ko ek ongoing habit banao, ek one-time checklist nahi. Rules evolve hote rahenge, transparency ki basic practice har situation mein safe rahegi.
Quick Verdict
Ghabrane ki zaroorat nahi hai, AI content banana legal hai. Bas ab chhupana allowed nahi hai. Label lagao, platform ka AI toggle use karo, aur kisi real insaan ka face ya voice bina permission ke recreate mat karo. Itna karo toh zyada tar creators safe hain.
Frequently Asked Questions
Kya AI se bana content India mein illegal hai?
Nahi, AI content banana aur post karna legal hai. Bas 20 February 2026 se ise clearly label karna legally zaroori hai.
Label kitna bada hona chahiye?
Law exact size fix nahi karta. Draft mein 10% screen area ka proposal tha lekin wo final rules mein shamil nahi hua. Ab label bas prominent aur easily noticeable hona chahiye.
Takedown ka time limit kya hai?
Court order ya government notice se flagged unlawful content ke liye roughly 3 ghante, jo pehle 36 ghante tha. Non-consensual intimate image complaints ke liye roughly 2 ghante.
Kya ye rules individual creators pe apply hote hain?
Direct obligations mainly platforms pe hain, lekin platforms ye requirements apne users tak pass karte hain, isliye practically creators ko bhi comply karna padta hai.
AI voiceover ke liye kya disclosure chahiye?
Synthetic audio ke liye shuruaat mein ek clearly prefixed spoken disclosure chahiye, matlab audio ke start mein batana ki ye AI-generated hai.
