Meta Platforms modified Instagram’s recommendation systems across India on Tuesday to curb the display of sexually suggestive and violent material, reporting that users now flag suggested posts about half as often as they did a year ago.
The adjustments alter what the platform actively promotes from accounts users do not follow, rather than instituting an outright ban on the underlying media.
The move follows an October 10 advisory from India’s technology ministry, which instructed the company to review its ranking engines, prevent the disproportionate amplification of manipulated or out-of-context content, and avoid trapping viewers within narrow narratives. Meta stated that default protections were upgraded over the past year using audience feedback, though it did not disclose specific underlying data.
The revised discovery model changes what Instagram distributes across the Explore tab, Reels, and algorithmic feed insertions. Unfollowed accounts producing borderline, suggestive, or aggressive posts face reduced distribution weights inside the recommendation architecture. Content remains accessible directly on creator profiles and to existing followers, preserving core publishing accounts while throttling viral reach.
New Delhi’s directive also mandated that Meta prevent the platform from funnelling users into repetitive, polarising content loops. The technology ministry instructed the US firm to audit how its recommendation engines cluster engagement around sensitive topics and report back on the technical mitigations deployed.
For consumer brands, digital retailers, and influencer agencies operating in India, the algorithmic adjustment raises customer acquisition costs on social channels. Organic audience expansion on Instagram relies heavily on non-follower recommendations through the Reels feed.
Source link







