Meta’s Algorithms, Audits And Safe Harbour Face Scrutiny In India

As Meta enters the second week of discussions with the government, the company's intermediary status and eligibility for safe-harbour protection have emerged as central issues in the talks. Officials are examining whether Meta's recommendation systems, paid-content promotion and other platform functions could go beyond the role traditionally associated with an intermediary under Indian law.

At the heart of the dispute is whether algorithms that determine what content users see could amount to publishing. Digital rights advocate Apar Gupta has disputed that interpretation, arguing that Section 79(2)(b) of the Information Technology Act refers to a human choice in selecting the receiver of a transmission, rather than automated ranking based on user activity and preferences. He has warned that treating recommendation systems as equivalent to content selection could have implications for search engines, spam filters, app stores and streaming services.

Independent tech-policy researcher Jyoti Panday, however, has placed the issue within the broader evolution of intermediary liability laws. She noted that safe-harbour protections were originally designed around platforms functioning as passive hosts, while questions arise when recommendation systems actively curate, prioritise and amplify unlawful material. She also pointed to US litigation involving TikTok, while acknowledging that the American legal framework differs from India's.

Also Read: MEA Clarifies No Government Role or Endorsement in Sheikh Hasina’s Private Press Event

Another major issue in the discussions is the government's demand for greater human intervention in content moderation. Officials want reviewers handling sensitive material to understand Indian languages, politics and local social context, particularly where automated systems may struggle to distinguish unlawful material from political satire or legitimate speech. Gupta argued that existing IT Rules allow some human oversight but do not require human review of every automated decision or provide a general power to demand algorithmic ranking weights or source code.

Deepfakes and takedown timelines have also become key points of contention. The government is seeking broader safeguards for manipulated and synthetic content, including stronger labelling and possible independent audits of algorithms. At the same time, officials have reiterated the three-hour timeline for specified unlawful-content takedown directions under the amended IT Rules. The discussions therefore involve a balance between rapid removal, contextual human review, due process and protection of legitimate speech.

The government is also considering whether similar questions should be examined in the case of other major platforms. Officials have indicated that the issue extends beyond Meta and involves determining whether platforms meet the intermediary definition and comply with the obligations attached to that status. Following two rounds of leadership-level meetings, Meta's India and technical teams are expected to continue discussions on moderation systems, human intervention, deepfake safeguards and technical controls around automated decisions.

Also Read: India And Japan Discuss Measures To Enhance Defence Cooperation