New Delhi (ABC Live): IT Rules 2026 for AI : India has taken its strongest regulatory step yet against deepfakes and AI-generated misinformation.

On February 10, 2026, the Ministry of Electronics and Information Technology notified the Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Amendment Rules, 2026. These rules will come into force on February 20, 2026.

At first glance, the government says the amendment responds to rising harms such as non-consensual intimate imagery, child sexual abuse material, and AI-driven impersonation fraud. However, the deeper impact is structural. In effect, India is turning digital platforms from passive hosts into active controllers of online speech.

This explainer explains what the rules change, why they matter, and how they could reshape free expression, platform responsibility, and India’s digital rights framework.

Why the IT Rules 2026 Matter

Today, deepfake technology is no longer experimental. Instead, it has become part of everyday digital life.

AI tools can now create realistic faces, voices, and videos in seconds. As a result, political propaganda, voice-clone scams, fake pornography, and false evidence are spreading fast.

Therefore, the State has moved away from a purely reactive approach. Rather than acting only after harm occurs, the 2026 Rules require platforms to stop, detect, label, and block synthetic content at the creation stage.

In simple terms, India is shifting from post-publication regulation to pre-publication control.

What Counts as “Synthetically Generated Information”

Under the amendment, synthetically generated information means:

Audio, visual, or audio-visual content that is artificially created or changed using a computer resource in a way that looks real and shows people or events in a way likely to be seen as real.

At the same time, the rules exclude:

• Routine editing and formatting
• Academic, educational, and research material
• Accessibility tools such as translation and transcription

Still, the main test is how real the content looks, not whether harm was intended.

Why This Matters

Because of this approach, content may fall under regulation even if it is:

• Satire
• Political parody
• Artistic recreation
• Documentary reconstruction

Importantly, the rules do not require proof of intent, deception, or harm.

Mandatory Labelling of AI-Generated Content

Under the new framework, platforms that allow creation or sharing of synthetic content must ensure:

• Clear and visible AI-generated labels
• Embedded permanent metadata or origin markers
• Unique identifiers linked to platform systems

Also, intermediaries must not allow removal or change of such labels or metadata.

Practical Impact

As a result, India is moving toward a nationwide AI-content watermark system.

Yet, the rules do not explain:

• What watermark format to use
• How labels will work across platforms
• How encrypted or offline content will be handled

So, platforms face high costs and legal uncertainty.

Platforms Must Prevent Illegal Synthetic Content

Under the amendment, intermediaries must use automated tools to ensure users cannot generate synthetic content that:

• Contains child sexual abuse material
• Creates false documents or electronic records
• Relates to explosives, arms, or ammunition
• Falsely shows real persons or real events

This is important, because it marks a big shift.

Platforms are no longer asked only to remove illegal content. Instead, they must stop it before it appears.

Three-Hour Takedown Rule

The amendment cuts timelines sharply:

• Takedown on lawful order: within 3 hours
• User response to grievance: 2 hours
• Grievance resolution: 7 days

Clearly, speed now drives enforcement.

Consequence

Because deadlines are so short, platforms will likely remove content quickly instead of checking it carefully. Therefore, over-removal becomes the safer choice.

Safe Harbour Becomes Conditional

The government says proactive removal using automated tools will not break Section 79 safe-harbour protection.

In practice, this means:

Safe harbour exists only if platforms actively monitor and moderate content.

Thus, intermediaries move from neutral hosts to mandatory content enforcers.

User Declarations and Platform Verification

Under the new rules, major social media platforms must:

• Ask users to declare whether content is synthetic
• Check the declaration with technical tools
• Show labels where synthetic content is confirmed

If, however, a platform knowingly allows unlabeled synthetic content, it is treated as failing due diligence.

Criminal Law Integration

Violations link directly to several criminal laws, including:

• Bharatiya Nyaya Sanhita, 2023
• Bharatiya Nagarik Suraksha Sanhita, 2023
• POCSO Act
• Representation of the People Act
• Indecent Representation of Women Act

As a result, AI-content moderation becomes part of criminal enforcement.

The Core Policy Shift

Simply put, India is moving from:

Notice and Takedown
➡ to
Detect, Label, Prevent, and Block

Taken together, this shows a preventive digital governance model.

Free Speech and Constitutional Concerns

Overbreadth

Because realism is the trigger, lawful speech can be regulated even without harm.

Chilling Effect

As a result, creators may self-censor.

Private Censorship

Therefore, platforms, not courts, decide what stays online.

No Appeals Framework

Meanwhile, users have no clear legal remedy against wrong takedowns.

How India IT Rules 2026 for AI Compares Globally

Overall, India follows the most intervention-heavy model among major democracies.

Bottom Line

The IT Rules 2026 for AI are not just about deepfakes.

Rather, they show India’s move toward a preventive digital state, where speech is filtered at creation and platforms act as delegated regulators.

Ultimately, courts and democratic debate will decide whether this path protects users or restricts free expression too far.

Also, Read ABC Live Report

Explained: How India’s Online Gaming Law Turns Criminal