IT (Intermediary Guidelines) Amendment Rules, 2026 | AI Act India
Official Gazette Notification · G.S.R. 120(E)
Ministry of Electronics and Information Technology · February 10, 2026

IT (Intermediary Guidelines & Digital Media Ethics Code) Amendment Rules, 2026

India's landmark regulatory update governing synthetic AI-generated content, deepfakes, and intermediary due-diligence obligations under the Information Technology Act, 2000. Effective 20 February 2026.

📋 Notification Number G.S.R. 120(E)
📅 Notified 10 February 2026
⚡ In Force 20 February 2026
🏛️ Authority MeitY · Section 87, IT Act 2000
📰 Gazette Issue No. 114 · Part II, Sec. 3(i)
🔍 Explore Key Changes 📄 Full Notification Text
Key Provisions

What the 2026 Amendment Introduces

🤖
Legal Definition of "Synthetically Generated Information"
For the first time, AI-generated deepfakes and algorithmically created audio-visual content receive a statutory definition — covering any content that appears indistinguishable from a real person or event.
New Clause (wa) · Rule 2
🏷️
Mandatory Labelling of AI Content
All platforms enabling AI-generated content must prominently label such content in visual displays, embed a unique permanent metadata identifier, and prevent its removal or suppression.
Rule 3(3)(a)(ii)
⏱️
Drastically Reduced Takedown Timelines
Content takedown timelines slashed from 36 hours → 3 hours. Grievance acknowledgement reduced from 24 hours → 2 hours. Complaint resolution from 15 days → 7 days.
Rule 3(1)(d) & 3(2)
🛡️
Safe Harbour Clarification for Compliant Platforms
Intermediaries that deploy reasonable technical measures and remove non-compliant content are explicitly protected from safe-harbour revocation under Section 79 of the IT Act.
Safe Harbour · Rule 2(1B)
📣
Enhanced User Notifications (Quarterly)
Intermediaries must inform users every three months about prohibited content categories, criminal liability for violations, and mandatory reporting obligations under BNSS 2023 and POCSO 2012.
Rule 3(1)(c)
📱
Significant Social Media: Pre-Upload Verification
Major social media platforms must require users to declare if content is AI-generated before upload and deploy technical tools to verify such declarations, displaying a clear label if confirmed.
New Rule 4(1A)
🔒
Prohibited AI-Generated Content Categories
Platforms must technically block AI-generated content involving CSAM, non-consensual intimate imagery, false documents/electronic records, explosive material instructions, and false deepfakes of real persons.
Rule 3(3)(a)(i)
👮
Police Authority Clarification
Police administration may designate multiple authorised officers — each at least at the rank of Deputy Inspector General — to issue content takedown orders via written mandate.
Rule 3(1)(d)(ii)(I)
⚖️
References Updated to BNS 2023
All references to the Indian Penal Code throughout the principal rules are substituted with the Bharatiya Nyaya Sanhita, 2023 (BNS), aligning IT Rules with India's updated criminal law framework.
Rule 7 Amendment
New Statutory Definitions

Rule 2 — Inserted Definitions

Clause (ca) · Rule 2(1)
Audio, Visual or Audio-Visual Information
Means any audio, image, photograph, graphic, video, moving visual recording, sound recording or other audio-visual content — with or without accompanying audio — whether created, generated, modified or altered through any computer resource.
Clause (wa) · Rule 2(1)
Synthetically Generated Information (SGI)
Audio, visual or audio-visual information artificially or algorithmically created, generated, modified or altered using a computer resource, in a manner that such information appears real, authentic or true — and depicts any individual or event indistinguishably from a natural person or real-world event.
Proviso (Exclusions): Routine editing, colour adjustment, noise reduction, transcription, PDF/document preparation, educational materials, accessibility improvements — without materially altering substance, context or meaning — are NOT deemed SGI.
Amendment Breakdown

Rule-by-Rule Changes

Rule 2 · Definitions
New Clauses (ca) and (wa) Inserted
Two new definitions inserted in Rule 2(1). New sub-rules 1A and 1B inserted after sub-rule (1):
  • Sub-rule 1A: References to "information" in rules 3(1)(b), 3(1)(d), 4(2) & 4(4) shall include synthetically generated information.
  • Sub-rule 1B: Removal or disabling of SGI content by a compliant intermediary does not violate Section 79(2)(a) or (b) of the IT Act — safe-harbour protection explicitly preserved.
Rule 3(1)(c) & (ca) — User Notifications
Quarterly Notices + New SGI-Specific Obligation
Clause (c) substituted — all intermediaries must now notify users at least once every 3 months of:
  • Their right to terminate/suspend access for non-compliance
  • User liability under IT Act and other laws for illegal content
  • Mandatory reporting obligations for cognisable offences (BNSS 2023, POCSO 2012)
New clause (ca): Intermediaries offering SGI-capable platforms must additionally warn users of penalties under BNS 2023, POCSO 2012, Representation of People Act 1951, Indecent Representation of Women Act 1986, POSH Act 2013, and Immoral Traffic (Prevention) Act 1956.
Rule 3(1)(d) — Government Orders
Takedown Time Reduced: 36 Hours → 3 Hours
Opening portion amended — response to government/court takedown orders must now occur within three hours (previously thirty-six hours). Orders from police administration must be in writing from an officer not below the rank of Deputy Inspector General of Police.
Rule 3(2) — Grievance Redressal
Faster Grievance Response Timelines
  • Complaint acknowledgement: 15 days → 7 days
  • First proviso response: 72 hours → 36 hours
  • Clause (b) response (urgent content): 24 hours → 2 hours
Rule 3(3) — NEW Sub-rule
Due Diligence for SGI-Capable Platforms
A wholly new sub-rule requiring SGI-capable intermediaries to:
  • Deploy automated tools to block SGI violating applicable laws — including CSAM, non-consensual intimate imagery, false documents, explosive material, and deceptive deepfakes of real persons.
  • Label all other SGI with prominent visual/audio disclosure and embed permanent, tamper-proof metadata with a unique identifier.
  • Prevent modification, suppression, or removal of labels or metadata by any user.
Rule 4(1A) — NEW Sub-rule
Significant Social Media Intermediaries: Pre-Upload SGI Declaration
Significant social media intermediaries must — before display, upload, or publication:
  • Require users to declare whether content is synthetically generated.
  • Deploy technical tools to verify the accuracy of such declarations.
  • Prominently label confirmed SGI content before it goes live.
Platforms that knowingly permit, promote, or fail to act on unlabelled SGI are deemed to have failed due diligence.
Rule 4(4) — Proactive Monitoring
"Endeavour to Deploy" Upgraded to Mandatory Obligation
Language change from "endeavour to deploy technology-based measures" to "deploy appropriate technical measures" — making proactive detection of previously removed content a firm compliance obligation rather than a best-efforts standard.
Rule 7 — Penalties
IPC References Replaced with BNS 2023
"Indian Penal Code" substituted with "Bharatiya Nyaya Sanhita, 2023 (45 of 2023)" — aligning the penalty regime with India's new unified criminal law.
Compliance Reference

Obligation Matrix — Who Must Do What

📋
Rule 2(1)(wa) New
Define & Recognise SGI in Policies
👥 All Intermediaries
📅 From 20 Feb 2026
🔔
Rule 3(1)(c) Amended
Quarterly User Notification on Prohibited Content & Liability
👥 All Intermediaries
🔄 Every 3 Months
⚠️
Rule 3(1)(ca) New
Notify Users of SGI-Specific Penalties & Consequences
🤖 SGI-Capable Platforms
🔄 Every 3 Months
🚨
Rule 3(1)(d) Amended
Takedown on Govt / Court Order
👥 All Intermediaries
⚡ Within 3 Hours
📬
Rule 3(2)(a)(i) Amended
Acknowledge Grievance Complaint
👥 All Intermediaries
⚡ Within 7 Days
🆘
Rule 3(2)(b) Amended
Urgent Content Grievance Response
👥 All Intermediaries
⚡ Within 2 Hours
🚫
Rule 3(3)(a)(i) New
Block Prohibited SGI — CSAM, Deepfake Fraud, False Records
🤖 SGI-Capable Platforms
🔁 Continuous / Automated
🏷️
Rule 3(3)(a)(ii) New
Label Permissible SGI + Embed Permanent Metadata & Unique ID
🤖 SGI-Capable Platforms
⬆️ At Time of Generation / Upload
📱
Rule 4(1A) New
Pre-Upload SGI Declaration from Users
📣 Significant Social Media Intermediaries
✅ Before Every Upload
🔍
Rule 4(4) Strengthened
Proactive Detection of Previously Removed Content
📣 Significant Social Media Intermediaries
🔁 Continuous Monitoring
Before vs. After

Tightened Response Time Requirements

🚨
Govt. / Court Takedown
3 hrs
36 hrs → Now
🆘
Urgent Grievance Response
2 hrs
24 hrs → Now
📬
Complaint Resolution
7 days
15 days → Now
⚖️
First Proviso Response
36 hrs
72 hrs → Now
🔔
User Notification Cycle
3 mo
Quarterly Mandatory
Official Record

Gazette Reference Details

Gazette Notification Number
G.S.R. 120(E)  /  सा.का.नि. 120(अ)
Document Reference
CG-DL-E-10022026-269993
Gazette Issue Number
No. 114 · Part II — Section 3 — Sub-section (i)
File Number
F. No. 2(5)/2025-CyberLaws
Signed By
Ajit Kumar, Joint Secretary, MeitY
Principal Rules
IT (Intermediary Guidelines & Digital Media Ethics Code) Rules, 2021
G.S.R. 139(E), 25 Feb 2021
Prior Amendments
G.S.R. 794(E) · 28 Oct 2022
G.S.R. 275(E) · 6 Apr 2023
G.S.R. 775(E) · 22 Oct 2025
Enabling Act
Information Technology Act, 2000 (21 of 2000)
Section 87 — Sub-sections (1) and (2)(z)(zg)
Date of Commencement
20 February 2026
Scroll to Top