How Deepfakes and AI-Generated CSAM in India Target Our Children

How Deepfakes and AI-Generated CSAM in India Target Our Children

Think of two scenes. A 13-year-old in Patna gets a friend request from a "schoolboy" who is actually a 45-year-old man. In Pune, a girl's class photo is lifted from Instagram, turned into a fake nude on a free app, and circulated on WhatsApp.

Neither scene needs a dark alley. Just a phone. An internet connection. And an adult with bad intentions.

This is AI-generated CSAM in India today. The law is now being tested by technology that can create fake sexual images in just a few seconds.

A few quick terms first

CSAM stands for child sexual abuse material. Our courts now prefer CSEAM "child sexual exploitative and abuse material." In September 2024, the Supreme Court said the term "child pornography" makes the abuse sound less serious. So, it told courts to use CSEAM instead.

A deepfake is simply a photo, video, or audio that is made by AI or changed by AI to look real.

Here is something most people miss: the image does not have to show a real child being abused to be illegal.

Section 2(1) (da) of the POCSO Act clearly covers a "digital or computer generated image indistinguishable from an actual child." It also covers an image "created, adapted, or modified" to look like a child.

So yes, sexually explicit AI made or AI changed pictures that seem to show a child can be punished under Indian law.

How grooming usually works

Grooming is when an adult slowly builds a friendship or trust with a child online. Their real goal is sexual exploitation.

It follows a pattern. A friendly stranger in a game or chat app. Daily messages. Compliments. Maybe a few game coins or a small phone recharge. Then a request for photos. Then threats "send more, or I show this to your family." This blackmail using sexual pictures is commonly called sextortion.

But the danger does not always come from a stranger.

NCRB's Crime in India data for 2024 shows a shocking number. In about 96.6% of cases under Sections 4 and 6 of POCSO which deal with penetrative sexual assault and aggravated penetrative sexual assault the offender was known to the child.

The biggest group? Friends, online friends, and live-in partners who promised marriage. Over 22,000 cases in just one year.

What the rulebook actually says

The law against AI-generated CSAM is wider than most parents think.

The POCSO Act, 2012 treats anyone under 18 as a child. Here is the breakdown:

Section 14: If someone uses a child for pornographic purposes, they get at least five years in jail and a fine for a first conviction. For a second conviction, the jail time goes up to at least seven years. If the person also sexually assaults the child, the punishment is much higher.

Section 15: This creates two separate offences.

  • Just keeping such material on your phone or computer without deleting, destroying, or reporting it even if you did not mean to share it means a fine of at least ₹5,000 for the first time. For a second offence, the fine is at least ₹10,000.
  • But if you keep it with the intent to share, spread, or display it, the punishment can be up to three years in jail, a fine, or both. Keeping it for a commercial purpose means three to five years for the first time and five to seven years if you do it again.

Sections 11 and 12: Even repeatedly contacting a child online with a sexual motive can be sexual harassment under this law. Similarly, threatening a child with sexual intent using a "real or fabricated" picture of their body can also attract POCSO. The punishment can go up to three years in jail plus a fine.

Read that again: a threat using a fake picture of a child is itself a crime under POCSO.

IT Act, Section 67B: This covers the online side. Creating, collecting, browsing, downloading, advertising, promoting, exchanging, or distributing sexual material involving children electronically can mean up to five years in jail and a fine of up to ₹10 lakh for the first time. Repeat offenders face up to seven years.

This section also covers some forms of online grooming like cultivating, enticing, or inducing a child into an online relationship for sexual activity. In simple words: the law does not have to wait until a sexual photo is actually shared.

The Supreme Court's clear message

On 23 September 2024, in the case Just Rights for Children Alliance v. S. Harish, the Supreme Court made an important ruling.

You cannot escape the law just because you did not permanently download a file. If you deliberately view such material online, it can amount to "constructive possession." That means you knowingly had control over it.

But this does not mean someone becomes a criminal by accidentally opening a wrong image. Knowledge, control, intention, and the specific crime involved still matter.

The new 2026 rules and the 3-hour deadline

Since 20 February 2026, India's updated IT Rules specifically address "synthetically generated information." This includes realistic AI-made and AI-tweaked content like deepfakes.

The rules now require online platforms to take extra steps against illegal synthetic content including child sexual exploitation material. They must also add clear labels and trackable metadata for AI-generated content where required.

There is also a stricter takedown timeline now. If a platform gets "actual knowledge" of illegal content through a court order or a valid government notice it must remove or block it within three hours. Earlier, the deadline was 36 hours.

To be clear: this does not mean every post reported by an ordinary user disappears within three hours. The three-hour clock starts only with a legally recognised order or intimation.

The messy reality of enforcement

Here is the uncomfortable truth.

NCRB's 2024 report recorded 69,191 POCSO cases. Cybercrimes against children? 1,238 cases were recorded. Out of these, 1,099 involved publishing or sharing sexually explicit material featuring children.

Remember these are only the recorded cases. They are not the full picture of what is actually happening online.

Enforcement does happen sometimes. In September 2022, the CBI's Operation Megh Chakra carried out searches at around 59 locations across 20 states and one union territory. It was a large operation against online sharing of child sexual abuse material.

But technology is running faster than the system. An image that once needed skills to edit can now be made in seconds. The law exists on paper. Making it work at internet speed is the harder challenge.

What you can actually do about it

The law usually steps in after the harm is done. Your job is to reduce the chances of it happening at all.

None of this costs money or needs technical skills.

Talk early and talk often. Do not give one big lecture. Give many small chats. Your child must know one thing above everything else: if something goes wrong online, you will not be angry at them.

Set simple rules. Never share personal photos, school name, home address, or live location with strangers. An online "friend" you have never met in person is still a stranger. If someone says "keep this chat a secret," that is a warning sign.

Use privacy settings. Keep social media accounts private. Limit who can send messages. Turn off chat with unknown players in games wherever possible.

Watch for warning signs. Sudden secrecy about the phone. A new older "friend." Unexplained gifts. Fear or nervousness after being online.

If the worst happens:

  • Do not panic. Do not scold.
  • Do not forward, repost, or copy the sexual image or video.
  • Save useful non-sexual evidence like usernames, profile details, website links, dates, and chat messages. If the offensive file is already on the device, report the matter and follow police instructions on how to keep it as evidence.

Where to report:

  • Go to cybercrime.gov.in. This is the national portal for reporting cybercrimes. It gives special attention to crimes against women and children.
  • Call the Child Helpline at 1098. Or use the POCSO e-Box on the NCPCR website.
  • Go to the police. Under Section 19 of POCSO, reporting certain offences is not a choice. An adult who does not report as required can face up to six months in jail, a fine, or both (Section 21). But a child cannot be punished under Section 21 for failing to report.

Indian law against AI-generated CSAM is stronger than most people realise. Fake images are not outside the law just because no real camera was involved.

But no law can replace a child who trusts you enough to say, "Something wrong happened online."

Build that trust today. It is the strongest firewall your family will ever have.

Categories:
Technology CrimeSexual HarassmentPrivacyOnline SafetyCyber Laws and PrivacyCriminal Law
Tags:
AI-generated CSAM in Indiaonline groomingPOCSO ActIT Act Section 67BdeepfakeCSEAMchild safety onlinecybercrime reporting India

Disclaimer

The Bar Council of India does not permit advertisement or solicitation by advocates in any form or manner. By accessing this website, www.advritujeph.in, you acknowledge and confirm that you are seeking information relating to Ritu Jeph Legal Services of your own accord and that there has been no form of solicitation, advertisement or inducement by Ritu Jeph Legal Services or its members.

The content of this website is for informational purposes only and should not be interpreted as soliciting or advertisement. No material/information provided on this website should be construed as legal advice. Ritu Jeph Legal Services shall not be liable for consequences of any action taken by relying on the material/information provided on this website. The contents of this website are the intellectual property of Ritu Jeph Legal Services.