Under accountability pressure in India, Meta for the first time commits to bypass a US intermediary and directly share child exploitation content reports with local law enforcement

Under continued pressure from India's federal government and the National Commission for Protection of Child Rights (NCPCR) over child sexual abuse material (CSAM), the US social media giant Meta disclosed several AI detection measures it has deployed in the Indian market, and for the first time stated that it would share relevant reports directly with Indian law enforcement, rather than relying solely on the previous single channel of the US-based National Center for Missing & Exploited Children (NCMEC).

According to the international edition of the Hindustan Times, Meta said in a press release that between January and June 2026, it took action on 5.3 million pieces of child sexual abuse material on its India-version Facebook and Instagram, of which more than 98% were proactively detected and actioned before being reported by anyone. Specific technologies include large language models for CSAM detection, enhanced advertising transparency, AI scanning of advertising content, "red team" AI agents used to discover system vulnerabilities, and the identification of offending accounts that re-register after being removed.

This disclosure came after a meeting between the Indian government and NCPCR with Meta. Earlier media reports had pointed out that CSAM content had appeared in advertising on the platform. A few weeks earlier, Meta had already said it would share them with Indian law enforcement in parallel with its reports to NCMEC.

However, the Hindustan Times report did not touch on several key links: the figure of 5.3 million pieces of content actioned was unilaterally disclosed by the company and was not verified by any independent third party; equally absent was Meta's action volume and proactive detection ratio in major European and American markets; the accuracy rate, false positive rate and impact on marginalized users of the newly deployed AI detection tools were likewise not disclosed by Meta.

The shift in reporting channels means that India has, for the first time, gained a direct route that does not pass through a US intermediary — in the past, Meta's reports of related violations mainly flowed to NCMEC, which is based in the United States, and law enforcement agencies in India and other countries typically had to obtain leads through that center. This mechanism has over many years been criticized as pointing to the lag in the accountability chain of Western tech giants when facing non-Western markets: in cases involving victims who are their own citizens, Indian law enforcement has sat at the downstream end of a cross-border chain. Meta's current commitment is a specific thaw under this long-standing asymmetric structure, but whether the commitment can be implemented in practical operational areas such as cross-border data transmission and data localization compliance remains limited in publicly available information.