Enhance monitoring: Social media needs stronger safeguards to protect kids

MeitY's action against Meta over CSAM ads highlights the need for stronger AI oversight, platform accountability and stricter safeguards for online child safety

The Ministry of Electronics and Information Technology’s (Meity’s) directive to Meta to disable all Instagram advertisements and content promoting or facilitating child sexual-abuse material (CSAM) highlights the urgent need to strengthen accountability of social-media platforms in an era of content distribution driven by artificial intelligence (AI). The fact that exploitative content reportedly appeared as paid advertisement raises deeper concerns not only about user-generated content but also about the integrity of automated advertisement-review systems and AI-based content moderation, which are supposed to undergo stricter scrutiny before publication. Meta has reiterated its “zero tolerance” policy and argued that it deploys advanced AI to detect abusive content across its billions of users. Yet, the controversy illustrates the limitations of AI-driven moderation. An investigation by the BBC showed that advertisements were being shown to sell such content though messaging app Telegram. It is reported that Telegram has removed over 274,000 groups and channels with such material this year. However, it doesn’t seem to be enough.

 

Recommendation engines and advertising algorithms are optimised to maximise engagement and revenue, not necessarily to detect sophisticated networks that constantly adapt to evade detection. When such content is monetised through advertisements, the failure extends beyond moderation to the platform’s commercial review mechanisms. This calls for independent scrutiny of both content-recommendation systems and advertising-approval processes rather than reliance on company assurances alone. The challenge is global and growing. In 2025, the United States-based non-profit organisation National Center for Missing & Exploited Children (NCMEC) received 21.3 million CyberTipline reports, including 61.8 million images, videos, and other files related to the child sexual exploitation. About 1.5 million reports had a generative-AI nexus, reflecting the rapid misuse of AI in child exploitation. The Internet Watch Foundation (IWF) also identified 8,029 AI-generated CSAM images and videos in 2025, a 14 per cent increase over the previous year, with a 260-fold surge in AI-generated videos.

 

The Information Technology Act and the Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Rules, 2021, impose obligations of due diligence on intermediaries. However, platforms that monetise content through advertising should face higher responsibilities. Independent algorithmic audits, transparent disclosure of advertisement-review processes, stronger know-your-advertiser (KYA) verification, rigorous monitoring of payment flows for suspicious advertising purchases and a mandatory human review of high-risk advertisements should become regulatory requirements. To further increase transparency, publishing periodic transparency reports by these platforms — detailing rejected advertisements, advertiser suspensions, and enforcement actions relating to child safety — will serve a useful purpose.

 

Equally important are meaningful deterrents. Repeated failures involving the paid promotion of illegal content should attract substantial financial penalties for non-compliance, while law-enforcement agencies must strengthen cross-border cooperation to dismantle transnational exploitation networks. AI-powered detection tools should complement trained human reviewers, particularly in high-risk areas such as child safety. The recently launched “Take It Down” initiative is also a step towards enabling the quicker removal of CSAM and non-consensual intimate imagery. As AI transforms the digital ecosystem, platform accountability must evolve from simply removing illegal content to preventing it from being financed, promoted, and amplified in the first place.