A disturbing question is sitting at the centre of India's latest social media debate: If platforms are getting better at detecting harmful content, how did advertisements allegedly promoting child abuse reach users in the first place?

The question has put Meta under intense scrutiny in India.

The Ministry of Electronics and Information Technology (MeitY) now says the company is actively removing Child Sexual Abuse Material (CSAM) and has become more cautious about such content. But the government is not calling the matter closed.

That is important. Because this story is no longer just about removing illegal posts. It is about how social media platforms moderate content, how advertisements are distributed, what recommendation systems do, and whether artificial intelligence is creating newer ways to exploit children online.

What Triggered the Child Sexual Abuse Material Controversy?

The controversy began after a BBC Eye investigation reportedly identified around 30 unique advertisements on Meta's platforms that allegedly promoted CSAM.

Some advertisements reportedly used phrases such as “rape video” and “child video” and directed users towards Telegram channels where illegal material was allegedly being offered.

The findings prompted MeitY to issue a notice to the company, directing it to disable advertisements and content that promoted or facilitated access to child sexual exploitative and abuse material.

The concern was particularly serious because these were reportedly paid advertisements. This meant the debate was not simply about an individual uploading illegal content. It also raised questions about how a platform's advertising systems allowed such material to reach users.

What Has Meta Done Since the Government Intervention?

According to MeitY officials quoted by Hindustan Times, the platform has become more cautious about unlawful content and has introduced additional filters to identify material that violates Indian law.

Its content moderators are also reportedly being made more sensitive to the Indian context.

The government has noticed a difference. Officials said they were no longer seeing the same kind of content emerging in the same way.

But there is an important catch.

MeitY has said it cannot yet declare itself fully satisfied.

That suggests the government is looking for evidence that the changes are working consistently, rather than treating a short-term reduction in harmful content as a permanent solution.

The company is also expected to provide details about how much content it has removed, how boosted content is targeted and how its moderation policies are being adapted for India.

Why Regional Languages Matter for Online Safety

One of the more significant demands reportedly made by MeitY concerns language.

Officials have asked for people who understand Indian languages beyond English and Hindi to help identify unlawful material on Instagram.

That may sound like an operational issue, but it is actually central to effective moderation.

India's internet is deeply multilingual. Harmful content can appear through slang, coded expressions, transliteration or regional languages that automated systems may struggle to interpret correctly.

A moderation system can only be as effective as its ability to understand context.

For platforms operating across India, that means technology needs to be supported by people who understand the country's linguistic diversity.

Is Removing Harmful Content Enough?

This is where the debate gets much bigger.

A social media platform does not simply host content anymore. Its systems can recommend posts, determine which advertisements reach particular audiences and influence how widely information travels.

That creates an important distinction:

There is a difference between failing to detect harmful content and having systems that help give that content greater visibility.

 

Technology policy experts quoted by NDTV have raised precisely this concern. The question is not necessarily whether a platform deliberately wants illegal material to spread. Instead, regulators need to examine whether its recommendation, advertising and ranking systems created foreseeable risks - and whether sufficient safeguards were in place.

This is why the current debate could have implications well beyond a single moderation incident.

What Does AI Change About the Problem?

Artificial intelligence has added another difficult layer to online safety.

Harmful material can now be created or manipulated synthetically, making traditional detection methods more challenging.

The issue became even more concerning after a separate investigation cited by NDTV reportedly found advertisements involving AI-generated abusive material and AI “nudify” applications.

Such tools can be used to generate fake explicit imagery from ordinary photographs, creating another avenue for image-based exploitation.

India has already moved to address this emerging challenge.

The Ministry of Electronics and Information Technology's 2026 amendments to the IT Rules specifically cover synthetically generated information. The rules require intermediaries that facilitate the creation or sharing of such material to deploy reasonable and appropriate technical measures to prevent unlawful synthetic content, including material containing child sexual exploitative and abuse material. They also introduce requirements around prominent labelling and provenance mechanisms for permitted synthetic content.

That means AI safety is no longer simply a future concern. It is already part of India's regulatory framework.

What Did Meta Tell the Government?

According to ThePrint, Meta told Indian officials that its systems were not fully equipped to handle the volume of unlawful content on its platforms and that it could not control all of it.

If accurately reported, that admission highlights one of the biggest challenges facing online platforms.

The nature of harmful content keeps changing. Users can create new accounts, use coded language, move conversations between platforms and use AI to generate material that traditional moderation systems may not recognise immediately.

This makes the government's focus on prevention, detection and accountability increasingly important.

ThePrint's report details the government meetings and the safe-harbour concerns raised during discussions with the company.

What Is Safe Harbour and Why Does It Matter?

Safe harbour is one of the most important legal issues in this controversy.

Under Section 79 of India's Information Technology Act, intermediaries can receive protection from liability for third-party content, provided they meet prescribed conditions and due-diligence obligations.

But that protection is not unconditional.

MeitY officials have reportedly told the company that safe harbour cannot be treated as a shield when the law is violated. At the same time, officials have clarified that the ministry does not simply grant or withdraw safe-harbour protection. The final legal determination would depend on the applicable legal process and, where necessary, the courts.

The debate becomes more complicated when a platform's own systems influence what users see.

If algorithms merely host content, one legal question arises. If they actively recommend, amplify or monetise illegal material, a different set of questions follows.

That is why the case could influence the wider conversation around platform accountability in India.

What Does Section 67B of the IT Act Say?

Section 67B of the Information Technology Act deals with electronic material depicting children in sexually explicit or indecent ways.

It covers activities including publishing, transmitting, creating, browsing, downloading and distributing prohibited material. It also addresses certain forms of online grooming and exploitation.

However, it is important not to confuse allegations with a legal finding.

The fact that harmful advertisements were reportedly discovered does not by itself mean the government has declared the company guilty under Section 67B. Investigations, evidence and judicial determination are separate steps.

Are Other Indian Authorities Investigating?

Yes.

According to NDTV, the National Commission for Protection of Child Rights (NCPCR) took suo motu cognisance of the allegations. The National Human Rights Commission (NHRC) also directed the Delhi Police to investigate the matter and examine whether relevant reporting obligations under the Protection of Children from Sexual Offences Act were followed.

The issue was also discussed during meetings between senior Indian officials and Meta executives in August.

NDTV reported that Mark Zuckerberg apologised over child abuse content, deepfakes and operational errors, citing sources familiar with the discussions. That remains a source-based report, rather than a publicly issued statement from Zuckerberg that has been independently verified in the material reviewed here.

So, Is the Problem Really Solved?

No - at least not yet.

There is evidence of corrective action. Government officials say Meta has become more cautious, strengthened filters and is removing problematic content.

But several questions remain unanswered.

How did the alleged advertisements get through the moderation process?

Can automated systems reliably detect new forms of AI-generated abuse?

Are advertising and recommendation systems creating additional risks?

Can moderation work effectively across India's many languages?

And how should intermediary protections apply when platforms themselves influence the visibility of content?

These are not questions that can be answered simply by counting takedowns.

The real test is whether platforms can prevent illegal and exploitative material from being promoted, recommended, targeted or monetised in the first place.

For users, that distinction matters.

Removing harmful content after it appears is necessary. But preventing it from gaining visibility is the bigger challenge.

MeitY's latest assessment suggests that Meta has taken steps in that direction. Yet the government's decision to continue monitoring the company shows that the issue is far from over.

And perhaps that is the most important takeaway: online safety cannot be measured only by what disappears from a platform. It must also be measured by what the platform prevents from reaching people in the first place.