Meta Takes Action on 5.3 Million Child Abuse Content Pieces, Launches New AI Ad Tools

Meta, the owner of Facebook and Instagram, announced on October 7 that it has implemented new artificial intelligence technologies to combat adverts that subtly direct people to content that abuses children. It took this action after disclosing that in the first half of 2026, it dealt with 5.3 million pieces of content in India that exploited children sexually. In a blog post, the business announced that they had discovered and removed over 98% of the content from Facebook and Instagram in India prior to any reports.

From January to June, it processed 33.2 million things across both platforms, with more than 97% of those items discovered proactively. According to the company, they are announcing new steps to further their fight against child sexual exploitation on its platforms today.

Meta Roles Out New Measures to Block Harmful Content

Meta has taken steps to combat these harmful methods by implementing new AI technologies and enhancing its systems. These adjustments are made to recognise adverts that either display child sexual abuse material or drive users to off-platform violating content. According to Meta, the continual effort to ensure the safety of its users' children is one of the company's top priorities. Moreover, it was pointed out that predators have started using ads as a weapon, since an ad can appear innocent at first glance but really directs users to off-site unlawful content.

 It expanded its probe beyond the reported ads after discovering the fraud. According to the firm, it also blocks the off-platform links and removes the accounts responsible. One of the new safeguards is a big language model system that can identify "signposting", or seemingly innocuous ads that are actually leading users to pirated or otherwise unlawful material. In addition to enhancing its assessment of an ad's path rather than just its display, Meta has used AI-driven sweeps to unearth content that was previously overlooked by previous systems.

Additionally, it has improved identification of banned users who create new accounts and uses a red-teaming AI bot to explore its own defences for vulnerabilities. In addition to this, Meta mentioned that it is still utilising image and video matching technologies like PhotoDNA, which it has integrated into its apps since 2011, as well as behavioural signals and systems based on rules.

Meta Aligning its Operations With India’s Norms

Meta made an announcement in September that it would submit incidents involving kid safety to the Indian Cybercrime Coordination Centre's National Cyber Crime Reporting Portal. Instead of targeting individual accounts, the business claims its investigators—some of whom have worked for the FBI—strike networks of predatory ones. According to Meta, the company is always cooperating with law enforcement in their efforts to bring those responsible for these crimes to justice. Concerned about the prevalence of deepfakes and other forms of child sexual exploitation on social media, India has increased its monitoring of these platforms.

Meta was sent with a notification by the authorities in July for content on Instagram that contained ads for the exploitation and abuse of children. Instagram was promptly instructed by the Ministry of Electronics and Information Technology (MeitY) to remove all adverts and content that promote or facilitate access to the explicit material. In addition, it called for a meeting of the company's top executives from around the world in India and asked for an exhaustive explanation. After many heated exchanges with Meta's upper management, the Ministry made it crystal apparent that the material in question is in breach of the law and is not eligible for safe-harbour protection.