Meta has changed how Instagram selects suggested posts in India, adding earlier AI screening for suggestive and violent material from accounts users do not follow in Explore, Reels and Feed.
Meta said the work took shape over the past year after feedback from users in India. Newer AI models now screen posts for recommendation eligibility before Instagram surfaces them to users, the company said. This moves the review earlier in the process than user reporting or hiding, according to Meta.
Suggestions can appear in Explore, Reels and Feed and may come from accounts a user has not chosen to follow. The India change concerns recommendations: Meta said it removes posts that violate its Community Standards and restricts content that is illegal in a given country.
Meta also identifies Sensitive Content Control and Not Interested as tools users can use to shape recommendations. The company said Sensitive Content Control adjusts suggestions within limits already applied in India.
Meta said people in India now report recommended posts at roughly half the frequency they did a year earlier. Its Nov. 16 announcement gave no count and did not explain the comparison method.
Separately, PTI reported, citing an Oct. 10 advisory it said it had seen, that India’s Ministry of Electronics and Information Technology asked Meta to review how its systems rank and recommend content. PTI said the advisory also called for context on material about ongoing public events and safeguards against disproportionate promotion of misleading, manipulated or out-of-context content.
Meta cited user feedback in its Nov. 16 announcement but did not mention the advisory or say whether it influenced the update. Reuters has also described earlier Meta efforts in Canada and Brazil to make Facebook surface less political material, a precedent for country-specific adjustments to content ranking.



