The platform cannot, cannot shirk its responsibility
In a digital age where commerce and conscience are perpetually in tension, a BBC investigation has revealed that Instagram's advertising infrastructure — the engine of a $200 billion empire — was approving and distributing paid promotions for child sexual abuse material across India, directing buyers to Telegram channels for as little as 80 pence. The platform's own moderation systems, when notified, initially found no violation. Only the weight of public exposure and government summons moved Meta to act — a sequence that raises enduring questions about whether profit-driven platforms can ever be trusted to police the darkest edges of human harm.
- Roughly thirty distinct paid ads using language like 'rape video' and 'child video' cleared Instagram's review process and reached users across India, some depicting children as young as twelve.
- Meta's automated moderation system actively rejected BBC abuse reports within 24 hours, declaring the content compliant with community guidelines — a failure that experts say reflects a structural indifference to safety over revenue.
- The Indian government summoned Meta representatives within hours of publication, and a retired Supreme Court justice warned the platform could not escape responsibility for profiting from criminal activity.
- A former Facebook vice-president who helped build the ad business called the findings 'horrifying but unsurprising,' describing an algorithm engineered for escalating engagement that inevitably produces exploitation as a byproduct.
- Experts and NGOs warn that criminals exploit the Instagram-to-Telegram pipeline to evade takedowns, and that dismantling organized trafficking networks will require cross-border intelligence cooperation that no single platform can provide.
Instagram's advertising system approved and circulated roughly thirty distinct ads promoting child sexual abuse material across India, a BBC investigation has found. The ads used explicit language and linked users to Telegram channels where material could be purchased for as little as 99 rupees — about 80 pence. Some content depicted children appearing to be around twelve years old, including one ad showing a man identified as 52 alongside a girl identified as 12. When the BBC reported these ads to Instagram, the platform's moderation team responded within 24 hours concluding the content did not violate community guidelines. Meta only disabled the ads and suspended the accounts after the investigation was published and the Indian government summoned company representatives to account for the failure.
The breakdown exposed deep structural problems in Meta's review process. The company claims every ad undergoes moderation — primarily automated, with human escalation for uncertain cases — yet the system failed repeatedly. A BBC test account in India, following women posting everyday content while using mild sexual innuendo, began receiving child exploitation ads within days. Meta later emphasized that no system is perfect, pointed to its proactive detection efforts, and noted it had disabled more than four million suspicious accounts in 2025. Telegram, meanwhile, claimed to have removed over 274,000 groups related to child abuse material in 2026, though one of two channels the BBC specifically reported continued posting new material for sale after the investigation.
Advertising accounts for nearly 98 percent of Meta's $200 billion annual revenue, and Instagram specifically derives more than 90 percent of its income from ads. In March, Meta announced it was reducing human moderation in favor of artificial intelligence. Brian Boland, a former Facebook vice-president who helped construct the advertising business, said he was 'horrified and unsurprised,' describing an algorithm designed to serve progressively more extreme content in pursuit of engagement — one that, without responsible oversight, would inevitably produce exploitation as an outcome. He left Meta believing the company 'didn't care about users anywhere.'
In India, 1.9 million reports of child sexual abuse material were received in 2025, second only to the United States. A Mumbai-based NGO noted that the vast majority of reports it receives originate from Meta platforms, and that criminals deliberately exploit the Instagram-to-Telegram pathway to evade moderation and reupload removed content. Experts stressed that the material is typically produced by organized criminal networks, including human traffickers, and that effective response will require international cooperation and intelligence sharing across borders — accountability that no single platform, left to its own commercial incentives, has yet demonstrated the will to provide.
Instagram's advertising system approved and distributed roughly thirty distinct advertisements promoting child sexual abuse material across India, a BBC investigation has revealed. The ads used explicit language—"rape video," "child video"—and directed users to Telegram channels where the material could be purchased for as little as 99 rupees, about 80 pence. Some depicted children who appeared to be around twelve years old. One showed a man identified as 52 with a girl identified as 12, inviting viewers to click through for more content. Another featured a very young girl in tears, with text suggesting sexual assault. When the BBC reported these ads to Instagram, the platform's moderation team responded within 24 hours saying the content did not violate community guidelines. Only after the investigation was published did Meta, Instagram's parent company, disable the ads and suspend the accounts behind them.
The Indian government moved quickly. Within hours of the BBC's publication, officials summoned Meta representatives to explain how such material had circulated on their platform. The discovery exposed a fundamental failure in Meta's advertising review process. The company claims every advertisement undergoes moderation before publication, relying primarily on automated technology designed to evaluate images, video, text, and audio, along with targeting parameters and destination links. When the system is uncertain, cases are escalated for human review. Yet the system failed repeatedly. The BBC set up a test account in India, following ten women who posted everyday content—food, weather, observations about daily life—while wearing revealing clothing and using sexual innuendo. Within a week, Instagram's algorithm began serving advertisements for video calls and explicit sexual content. Days later came the child exploitation ads.
Meta's response emphasized that no system is perfect and that its review process may not detect all violations. The company noted it runs proactive detection on live ads and that anyone can report content they believe breaks the rules. When it becomes aware of apparent child exploitation, Meta said it reports the material to the National Center for Missing and Exploited Children, the centralized global reporting system for online sexual exploitation of children. The company also stated it had disabled more than four million accounts in 2025 showing signals of suspicious behavior and was constantly developing new technology to identify predators and block links to violating websites.
Telegram, the messaging platform where the material was being sold, told the BBC it had removed more than 274,000 groups and channels related to child sexual abuse material in 2026. The company uses both automated and human moderation and claims to have "virtually eliminated" the public spread of such material from its platform. However, when the BBC reported two specific channels selling child abuse videos, one was taken down but the other continued posting new material for sale. Telegram is not a member of either the NCMEC or the Internet Watch Foundation, organizations that work with most platforms to find and remove such content.
Advertising is central to Meta's business. In January, the company reported that nearly 98 percent of its $200 billion revenue for 2025 came from advertising. Analysts estimate ads account for more than 90 percent of Instagram's revenue specifically. In March, Meta announced it was reducing reliance on third-party human moderators and increasing the use of artificial intelligence, saying experts would design, train, oversee, and evaluate these systems. Brian Boland, a former Facebook vice-president who worked at the company between 2009 and 2020 and helped build the advertising business, said he was "horrified and unsurprised" by the findings. He described Instagram's algorithm as designed to keep users engaged by showing them "something more extreme, more tantalising," and said that while the algorithm wasn't deliberately creating pedophiles, the company's failure to responsibly control it in pursuit of revenue and clicks would inevitably produce such outcomes. Boland left Meta because he believed the company "didn't care about users anywhere" and deleted his Instagram account in 2025.
A retired justice of India's Supreme Court, Madan Lokur, expressed alarm that Instagram was "making money by participating in a criminal activity." He said the issue was serious enough for India's highest court to take suo moto cognisance—initiating legal proceedings without waiting for a case to be brought—and that despite Indian law protecting social media companies from liability for user-uploaded content, "the platform cannot, cannot shirk its responsibility." Boland testified against Meta in a trial in New Mexico earlier this year in which the company was accused of misleading users about platform safety for children. A court ordered Meta to pay $375 million to the state; Meta disagreed with the verdict and said it intended to appeal.
In India, 1.9 million reports of child sexual abuse material were received in 2025, second only to the United States with two million. Shikha Goel, director of the Cyber Security Bureau in the Indian state of Telangana and one of the country's top cyber police officers, said Instagram and Facebook generated the most reports but cautioned that this reflected their detection capabilities rather than necessarily indicating they were the largest sources. A Mumbai-based NGO called the Rati Foundation, which runs a helpline for children facing online harms, said the vast majority of reports it receives come from Meta platforms. Its co-founder, Siddharth Pillai, noted that "criminals use the seamless navigation from Instagram to Telegram to evade our moderation efforts, and keep reuploading the content we help take down." Experts said child sexual abuse material in India was typically created by organized criminal groups, including human traffickers, though family and community members were sometimes involved. Bhuwan Ribhu, founder of Just Rights for Children, a network of more than 250 organizations working to prevent violence against children in India, said the crime remained underreported and police were still developing the technical skills to address it effectively. He emphasized that international cooperation and intelligence sharing across borders would be vital to track the full chain of demand and supply in organized trafficking networks.
Citações Notáveis
This is a serious enough issue for the Supreme Court of India to take suo moto cognisance and get the government to act against any social media platform.— Madan Lokur, retired justice of India's Supreme Court
Criminals use the seamless navigation from Instagram to Telegram to evade our moderation efforts, and keep reuploading the content we help take down.— Siddharth Pillai, co-founder of the Rati Foundation
A Conversa do Hearth Outra perspectiva sobre a história
How does an ad promoting child abuse material even get approved by Instagram's system in the first place?
The company relies primarily on automated technology to review ads before they go live. The system checks images, video, text, audio, targeting, and destination links. When it's uncertain, cases go to human reviewers. But the BBC found the system failed repeatedly—it even rejected their own reports, saying the content didn't violate guidelines.
So the algorithm itself is broken, or is it something else?
It's more complicated than a broken algorithm. A former Facebook vice-president said the algorithm is designed to keep people engaged by showing them increasingly extreme content. The problem isn't that it's trying to create exploitation—it's that Meta isn't responsibly controlling it because the priority is revenue and engagement, not safety.
But Meta says they're fighting this aggressively. They disabled accounts, they report to authorities.
They do those things, yes. But the scale matters. In 2025, India received 1.9 million reports of child abuse material—second only to the US. Meta platforms generated the most reports. And when the BBC tested the system by creating a single account, within days it was being served ads depicting children in sexual situations.
Why is Telegram part of this at all?
The ads link directly to Telegram channels where the material is sold for about 80 pence. Telegram removed some channels when the BBC reported them, but others kept operating. The company isn't part of the major reporting networks that most platforms use to coordinate on this issue.
What would actually change this?
A former Meta executive said if people en masse deleted their accounts, the company would pay attention. But experts also emphasize that this requires international cooperation—tracking organized trafficking networks means following the entire supply chain across borders, which no single platform can do alone.
Is there any accountability happening?
The Indian government summoned Meta. A court in New Mexico ordered Meta to pay $375 million for misleading users about child safety, though Meta is appealing. But the fundamental issue—that advertising revenue incentivizes the company to move fast and break things, even when those things are children—remains unresolved.