Meta is facing severe backlash after its automated moderation systems approved and displayed paid advertisements promoting child sexual abuse material (CSAM) to Instagram users in India. A recent investigation by BBC Eye revealed that the platform's ad network actively monetized and served dozens of illegal advertisements, exposing a critical failure in how social media giants police paid content.
The investigation uncovered approximately 30 unique advertisements promoting CSAM, alongside 20 ads featuring adult pornography. These paid promotions utilized explicit terms like "rape video" and "child video," directing users to Telegram channels where illegal content was allegedly sold for as little as 99 Indian rupees, or roughly $1. To test the algorithm, researchers created a fresh Instagram account that followed ordinary but suggestive content.
Within days, the platform's recommendation engine escalated from showing adult pornography to serving ads depicting children in sexually suggestive situations. The most alarming failure occurred when researchers reported an advertisement showing a distressed young girl accompanied by text implying sexual assault. Twenty-four hours later, Instagram's automated system responded that the ad did not violate its community guidelines and allowed it to remain active.
Meta only disabled the ads, suspended the associated accounts, and blocked the related URLs after being directly contacted by the publication for comment. A Meta spokesperson acknowledged that "no system is perfect" and stated the company is continuously improving its automated detection technologies while reporting apparent cases to the National Center for Missing and Exploited Children (NCMEC).
The Role of Telegram and Algorithmic Amplification
The distribution network relied heavily on Telegram to finalize the illegal transactions. While one reported Telegram channel was removed, another continued to upload new content. Telegram defended its practices, claiming it utilizes both automated systems and human moderators, resulting in the removal of over 274,000 groups and channels linked to CSAM in 2026.
However, former Facebook executive Brian Boland noted he was "horrified and unsurprised" by the findings. He explained that recommendation systems inherently prioritize engagement and revenue, warning that without stronger safeguards, these algorithms can unintentionally amplify increasingly extreme content.
The scale of this crisis is particularly severe in India, which recorded 1.9 million CyberTipline reports related to CSAM in 2025 - second only to the United States. Retired Indian Supreme Court Justice Madan Lokur emphasized that social media platforms cannot evade responsibility simply because users upload the illegal content, suggesting the issue warrants intervention from India's Supreme Court.
The Financial Incentive Blind Spot
The fact that these were paid advertisements rather than organic posts highlights a dangerous blind spot in Meta's moderation pipeline. While platforms heavily scrutinize organic viral content, ad-approval algorithms are often optimized for frictionless onboarding to maximize revenue. When bad actors can literally pay Meta to distribute CSAM, the "no system is perfect" defense falls flat.
This incident will likely accelerate regulatory action in both India and the European Union. With India generating nearly 2 million CyberTipline reports annually, local lawmakers have concrete data to force Meta and Telegram into implementing mandatory human-in-the-loop reviews for specific ad categories. Until financial penalties exceed the ad revenue generated by these automated systems, algorithmic amplification of illegal content will remain a persistent threat.
