NEW DELHI: Meta’s platforms must operate in accordance with Indian law rather than relying solely on the company’s global policies, government sources said on Sunday, while clarifying that it would be inaccurate to suggest that the Centre had directed the technology major to completely overhaul its algorithm.
Officials said the primary objective of recent engagements with Meta was to ensure that its platform guidelines, moderation systems and content policies remain fully aligned with Indian laws and regulatory requirements.
Sources said India’s cultural and linguistic context is often not adequately understood by teams operating outside the country. They stressed the need for greater local input, including a stronger understanding of Indian languages, social sensitivities and cultural nuances, to improve content moderation.
On Child Sexual Abuse Material (CSAM), officials maintained that there could be no tolerance or compromise. They said allowing such content to remain accessible on platforms would constitute a serious violation of rules and called upon Meta to adopt stronger and more proactive measures.
The government, sources said, expects concrete action rather than assurances alone to prevent the circulation and reappearance of such material.
Officials also raised concerns over the handling of deepfakes involving prominent personalities. They said content uploaded from authorised or verified handles should not be wrongly classified as deepfake material.
Such sensitive cases, they added, should not be left entirely to automated detection systems. The government has called for a “human-in-the-loop” mechanism to ensure that appropriate review takes place before any takedown action is initiated.
Concerns were also expressed over synthetically generated content resurfacing after it had already been identified and removed as a deepfake. Meta has been asked to explain the concrete measures being taken to prevent previously removed content from repeatedly reappearing on its platforms.
The government has further sought clarity over Meta’s recommendation and virality algorithms, questioning why users are frequently served content unrelated to their stated interests.
Sources said this appeared to indicate shortcomings in the recommendation system that required examination. Officials also questioned why synthetically generated content continued to receive algorithmic amplification even after being flagged.
They stressed that once such material has been identified, it should neither be recommended further nor repeatedly resurface on the platform.
The clarification followed a meeting on Friday between Meta’s technical team and officials of the Ministry of Electronics and Information Technology (MeitY), during which company representatives explained existing systems for dealing with harmful online content.
Technical consultations between the government and Meta are continuing, and another round of discussions is expected in the coming week.
