New Advisory Requires Meta to Guard Against Algorithmic Bias in India

On October 19, the Ministry of Electronics and Information Technology issued a new advisory that compels Meta to reassess the bias potential in its Facebook and Instagram recommendation engines. The guidance stresses that the platforms must avoid presenting users with a single ideological stream repeatedly.

Meta is urged to recalibrate its systems to surface a multiplicity of viewpoints, ensuring that information remains diverse, accurate, and credible. The advisory also calls for the deployment of enhanced monitoring tools to track content that spreads quickly through algorithmic promotion and could disrupt public peace.

The ministry specifically points to visual content—photos, videos, graphics—related to public incidents that may suffer from ambiguous provenance. It demands that any such material be accompanied by verifiable details about its origin, timing, and context.

Instances where outdated or unrelated media are re‑used as evidence of current events, where clips have been selectively edited, or where unverified claims about police or government actions are made must be flagged and contextualized.

This directive follows previous engagements with Meta in August, when officials questioned the company about safeguards against deepfakes, synthetic media, and prohibited content. It underscores the government's intensified focus on algorithmic transparency and responsible content curation.