India tells Meta to add human review to AI content moderation systems

Indian authorities have flagged concerns about Meta's reliance on algorithmic content moderation, particularly the use of keyword-based systems that can remove posts without adequate contextual analysis. The government has instructed the social media giant to incorporate human review mechanisms into its AI-led takedown processes to prevent the removal of legitimate content.

The directive addresses a systemic issue where automated systems identify flagged keywords or phrases and remove associated posts without considering the broader context in which they appear. This approach has led to instances where legitimate discussions, news reporting, or culturally significant content gets removed, raising questions about content moderation standards and free expression online.

The intervention reflects growing scrutiny of how technology platforms moderate content in India, where regulatory bodies have increasingly engaged with Meta and other social media companies over content policies. The government's position emphasizes the need for a balanced approach that combines artificial intelligence efficiency with human judgment to ensure fair and contextually appropriate moderation decisions.

Meta has been directed to review its current moderation framework and implement changes that prioritize human verification, particularly for sensitive categories of content. The move underscores India's broader effort to ensure that content moderation practices align with local regulations and respect legitimate speech while addressing genuinely harmful material.