Skip to content
VECTOR WIREAI INTELLIGENCE
UTC
Home

Meta retools Instagram recommendation AI for India to curb sexual, violent content

Indian users report recommended posts about half as often as a year ago; absolute totals not disclosed.

Meta has modified Instagram's recommendation systems specifically for India to reduce the visibility of sexually suggestive and violent material surfaced from accounts users do not follow1,2. The changes, announced on November 16, apply to content recommended across Explore, Reels, and Feed4.

Meta said the measures concern what Instagram's algorithms promote rather than a blanket ban on such content being posted on the platform3. The company said it strengthened Instagram's default content safeguards over the past year, drawing on feedback from users in India.

Upgraded detection models

The core technical shift: Meta's systems now use more advanced AI models to assess posts against Instagram's recommendation guidelines before surfacing them, rather than relying primarily on users to report or hide content. The company said the upgraded detection systems have been deployed globally, while the tighter recommendation filters are India-specific.

Meta said Instagram users in India now report recommended content about half as often as they did a year ago5. The company did not disclose the absolute number of reports or how much content it had filtered out.

Instagram users can further limit sensitive content through Sensitive Content Control, which covers Search, Reels, Explore, Feed recommendations, and suggested accounts. A separate Political Content Control lets users see less political content from accounts they do not follow.

Regulatory backdrop

The update follows pressure from New Delhi, which called on platforms to review how algorithms amplify misleading, violent, and sexually suggestive material. India's government has been pressing social platforms not just to remove illegal content but to rein in how recommendation systems distribute it.

The move also comes amid scrutiny over the alleged circulation and promotion of child sexual abuse material on Instagram; a BBC investigation alleged that paid Instagram ads directed users to Telegram channels offering such material. Meta faces broader demands from Indian authorities over harmful content and the potential misuse of features across its platforms6.

ANALYSIS The shift from reactive moderation — waiting for user reports — to pre-recommendation AI screening marks a change in how Meta allocates enforcement resources for the Indian market, placing the filtering burden on its own models rather than on users.

Was this useful?

The day’s vector, in your inbox.

Vector Wire Daily: one dispatch each publishing day.