Instagram Under Scrutiny After Child Abuse Advertisements Reach Users

Instagram presents itself as a place for entertainment, creativity and social connection. Yet a recent investigation has raised disturbing questions about what the platform recommends, what it allows advertisers to promote and how quickly harmful material can reach ordinary users.

The BBC investigation, published in early July 2026, reported that paid advertisements on Instagram appeared to promote child sexual exploitation and abuse material. Some of these advertisements allegedly directed users to Telegram channels where illegal videos were offered for sale.

The findings have placed Meta, the company that owns Instagram, under intense pressure in India, one of its largest and most important markets.

How the investigation began

The BBC created a fresh Instagram account on a new device to examine how the platform’s recommendation system behaved.

The account followed female creators who posted revealing or sexually suggestive material. According to the investigation, Instagram soon began recommending increasingly explicit content.

Within days, the feed reportedly started displaying advertisements containing nudity and sexual activity, despite Meta’s advertising rules prohibiting such material.

The content then became significantly more disturbing.

The investigation said advertisements began appearing in which children were sexualised or presented in abusive situations. Some included coded captions offering additional videos through private messages.

Instead of remaining inside Instagram, the advertisements acted as entry points to Telegram channels where illegal material was allegedly being sold.

The BBC investigation suggested that this process did not require the account to search directly for child abuse material. The recommendation and advertising systems reportedly moved the account towards it after detecting engagement with other sexualised content.

The original transcript also states that when one advertisement involving a child was reported to Instagram, the platform’s system initially responded that the advertisement did not violate its community guidelines.

Why paid advertisements make the case more serious

Harmful content uploaded by an individual user is already a serious moderation failure. Harmful content distributed as a paid advertisement raises additional concerns.

An advertiser pays the platform to place material in front of selected audiences. Meta has previously said that advertisements are reviewed before publication. An advertisement appearing in a user’s feed therefore suggests that it passed through, avoided or manipulated the company’s review process.

Previous investigations have identified similar weaknesses. In 2025, AI Forensics found thousands of explicit advertisements running on Facebook and Instagram. Researchers reported that the same images were removed quickly when uploaded as ordinary posts, while paid versions remained active. This raised questions about whether paid content receives different treatment from regular user content.

The latest case is far more serious because it concerns the alleged commercial distribution of child sexual abuse material.

It is important to distinguish between consensual adult content and material involving children. Adult sexual content may violate platform rules or national obscenity laws. Child sexual abuse material is evidence of abuse. Producing, possessing, promoting or selling it is a serious criminal offence.

Instagram as the entrance to a wider criminal market

According to the investigation, Instagram was not always the final destination. It was allegedly being used as a public storefront.

Advertisements attracted attention on Instagram and then directed interested users to private Telegram channels. These channels reportedly offered collections of illegal material at low prices.

This method gives criminal networks several advantages.

Instagram provides access to a massive audience and a powerful recommendation system. Telegram provides private groups, channels and direct communication. Payments and file transfers can then take place away from the original advertisement.

The Financial Times reported that the BBC investigation found advertisements and content promoting child sexual exploitation, with some links leading to Telegram channels where illegal material was offered for purchase.

The movement of users from a public social network to a private messaging service also makes investigations more difficult. Accounts can disappear, channels can be renamed and sellers may operate outside the country where victims or buyers are located.

How recommendation systems can create a harmful path

Instagram does not show every user the same feed.

Its recommendation system studies behaviour and predicts what might keep each person watching. Even without a direct search, watching a video, opening a profile or pausing on a post can influence future recommendations.

This system is designed to increase engagement. The longer people remain inside the application, the more opportunities the platform has to show advertisements.

The exact formula used by Instagram is not public. However, regulators increasingly argue that personalised recommendations, automatic video playback, endless scrolling and repeated content loops can encourage compulsive use.

In July 2026, the European Commission accused Meta of failing to address risks created by features including infinite scrolling, automatic playback and engagement focused recommendations. European regulators said these designs may be particularly harmful to children and warned that Meta could face major financial penalties if it fails to make changes. Meta rejected the criticism and pointed to its existing safety measures.

The danger is not simply that harmful material exists. It is that a recommendation system can repeatedly place similar material in front of someone once it detects a sign of interest.

A user may begin with entertainment or suggestive content and gradually receive stronger material. The system does not need to understand the moral or legal meaning of every video. It only needs to predict whether the person is likely to continue watching.

The attention economy behind sexualised content

For many creators, attention can be converted into money.

A video that receives large numbers of views may bring followers, subscriptions, brand partnerships or private paying customers. Sexualised content often attracts quick reactions because it creates curiosity, surprise or controversy.

Some creators use Instagram subscriptions to offer exclusive posts and direct interaction. Others use Instagram to send followers to private groups or external websites.

These activities are not automatically illegal. The problem begins when creators violate platform rules, deceive audiences, target children or redirect users towards abusive and criminal material.

Criminal networks can use the same attention economy. They may post coded advertisements, censored previews or misleading captions that avoid automatic detection while making the intended message clear to potential buyers.

The platform’s commercial interests can create a conflict. Content that generates strong engagement is valuable to an advertising business, but provocative material may also be harmful. Effective safety systems must therefore be strong enough to restrict content even when that content attracts attention.

Why children face greater risks

Children and teenagers are especially vulnerable to recommendation systems.

Their ability to assess risk and control impulses is still developing. They may also be more curious, more influenced by social approval and less prepared to recognise grooming or manipulation.

A 2025 study examined 3,000 videos recommended by Instagram, TikTok and YouTube to accounts assigned ages of 13 and 18. Researchers found significant weaknesses in moderation and reported that harmful material could reach younger users without them actively searching for it. The study also found that age restrictions were easy to bypass by entering a false age.

Exposure to sexualised content does not affect every child in the same way. However, repeated exposure can distort expectations about relationships, appearance, consent and personal boundaries.

The greatest concern is not only what children may see. It is also who may find them.

Platforms that recommend accounts to one another can unintentionally connect vulnerable young users with adults seeking to exploit them. This risk becomes greater when age verification is weak and private messaging remains easily available.

The problem with automated moderation

Instagram receives an enormous quantity of videos, images, messages and advertisements every day. Reviewing every item manually would be extremely difficult.

Meta and other technology companies therefore use artificial intelligence, automated filters, digital fingerprints and human reviewers.

These systems can identify known illegal images and common patterns. They are less effective when criminals use new material, coded language, altered images, hidden frames or misleading captions.

Context is another major challenge.

A machine may identify exposed skin but fail to understand whether a child is being sexualised. It may detect certain words while missing slang or coded expressions. It may also make mistakes when reviewing content in languages for which fewer moderators and training resources are available.

This explains how platforms can sometimes remove harmless posts while failing to stop clearly dangerous material.

It does not remove the company’s responsibility. A platform that earns money from targeted advertising must also invest enough resources to ensure that criminal advertisements are not approved, distributed or monetised.

Meta’s response

Meta has said it has zero tolerance for child sexual exploitation and abuse material.

Following the investigation, the company said its systems had already identified and disabled several violating advertisements and the accounts behind them. It said additional advertisements were removed after the BBC raised the issue.

Meta disputed any suggestion that it knowingly or deliberately targeted users with illegal advertisements.

The company’s response confirms that violating advertisements and accounts existed. The key questions are how they were approved, how widely they were shown, how much money was collected from them and why the reporting system did not stop them earlier.

A zero tolerance policy is meaningful only when enforcement works consistently.

India orders Meta to act

The Indian government responded shortly after the investigation became public.

The Ministry of Electronics and Information Technology directed Meta to immediately disable advertisements and content that promoted or facilitated access to child sexual exploitation and abuse material. The company was also asked to explain how the content remained active and what corrective measures it was taking.

Meta was given seven days to provide an explanation.

By July 13, 2026, Meta had submitted its response. India’s Information Technology Secretary said the reply was being examined and that further action could follow after the review.

The government later reminded social media companies that they must follow procedures for identifying and reporting child sexual abuse material and cooperate with law enforcement agencies, including the Indian Cyber Crime Coordination Centre.

What Indian law says

India already has strong legal provisions covering this type of material.

Section 67 of the Information Technology Act deals with obscene material published or transmitted electronically. Section 67A covers material containing sexually explicit acts.

Section 67B specifically covers material depicting children in sexually explicit situations. It applies not only to publishing or transmitting such content, but also to collecting, seeking, browsing, downloading, advertising, promoting, exchanging or distributing it.

The Protection of Children from Sexual Offences Act also criminalises the use of children for sexual material. Section 15 provides punishment for possessing, storing, transmitting, displaying, distributing or commercially using such material, subject to the circumstances and intention involved.

Social media companies may receive limited protection from liability for content uploaded by third parties. Under section 79 of the Information Technology Act, that protection depends on the intermediary observing due diligence and acting promptly when it obtains knowledge of unlawful material.

The central issue is therefore not the absence of law. It is enforcement, technical capacity, international cooperation and platform accountability.

What other countries are doing

Governments around the world are experimenting with stricter rules for social media companies.

Australia introduced a minimum social media age of 16, requiring major platforms to take reasonable steps to prevent younger children from holding accounts. The rules took effect in December 2025 and allow penalties of up to 49.5 million Australian dollars for serious failures.

Platforms later reported removing millions of accounts believed to belong to users younger than 16. However, questions remain about privacy, accuracy and the ability of children to bypass age checks.

The European Union is taking a different approach. Rather than relying only on age restrictions, it is examining the design of recommendation systems, endless feeds and addictive features. Regulators can demand risk assessments, independent audits and changes to platform design.

These approaches show that bans alone are unlikely to solve the problem. Children may move to smaller and less regulated services. Strict identity checks can also create privacy risks.

A stronger model would combine reliable age assurance, independent algorithm audits, rapid reporting systems, human review, advertiser verification and major penalties for repeated failures.

What Meta and other platforms should change

First, advertisements involving children or sexual themes should face specialised human review before publication.

Second, advertisers should undergo stronger identity and payment verification. Platforms must be able to identify who purchased an advertisement and where that person or organisation is located.

Third, child safety complaints should be reviewed by trained specialists rather than handled only through automatic systems.

Fourth, platforms should examine the full journey created by an advertisement. Removing a single post is not enough when the account, payment method, related pages and external links remain active.

Fifth, independent researchers and regulators should receive access to meaningful data about how recommendations are produced and how many users are exposed to harmful advertisements.

Finally, platforms should be required to disclose whether they earned money from advertisements later found to be illegal and whether that money was returned, frozen or handed to authorities.

A test of responsibility

The investigation is not simply about a few disturbing advertisements.

It raises a larger question about whether a platform can claim to be safe when its own systems approve, distribute and earn money from material that violates both its policies and the law.

Meta says it does not tolerate child sexual exploitation. Indian authorities are now examining whether the company’s actions match that promise.

The final responsibility cannot be placed on users, parents or children. A person should not have to understand a complex recommendation system to avoid being exposed to criminal material.

Instagram designed the feed, operates the advertising system and controls the recommendation technology. That gives the company enormous influence.

It also gives the company an unavoidable duty to protect the people using it.