India Pushes Meta for Stronger Deepfake Checks, Greater Accountability on Content

Photo: IANS

The government has stepped up its engagement with Meta over the way harmful and misleading content is handled on its platforms, asking the company to ensure that its content policies and enforcement practices are in line with Indian laws and the country’s social and cultural context.

The government has, however, made it clear that it is not seeking a fundamental overhaul of Meta’s algorithms. Instead, the focus is on improving safeguards, strengthening compliance and making the company’s content moderation and recommendation systems more transparent.

According to an NDTV Profit report, discussions between the government and Meta are likely to continue, with deepfake content emerging as one of the key issues. Authorities have reportedly asked the company to introduce a “human-in-the-loop” system for reviewing suspected deepfakes, so that potentially sensitive takedown decisions are not left entirely to automated tools.

The move comes amid growing concerns over the ability of AI-generated videos, images and audio to convincingly imitate real people and events. While automated detection can identify a large volume of suspicious material, the government believes human review is important in cases where context is difficult for an algorithm to understand. Such oversight could also help prevent genuine content from being wrongly removed.

Officials have reportedly told Meta that content posted by prominent verified accounts should not be automatically treated as deepfake material. Instead, such content should be subjected to appropriate human scrutiny before any enforcement action is taken.

The government has also asked Meta to ensure that deepfake content removed from its platforms does not simply return through repeated uploads. This would require stronger systems for identifying duplicate or modified versions of previously removed material and limiting its continued circulation.

Another issue receiving priority is child sexual abuse material, or CSAM. Authorities have reiterated their zero-tolerance approach and stressed that platforms must have effective mechanisms to detect, remove and act against such content without compromise.

Beyond individual pieces of content, the government is also seeking greater clarity on how Meta’s recommendation systems operate. Officials want more transparency about how posts are selected, ranked and presented to users, particularly when those systems can significantly influence what people see and engage with online.

The broader concern is that content moderation is no longer simply about removing clearly illegal material. Recommendation algorithms can amplify misleading or harmful content at a much larger scale, making transparency and accountability increasingly important.

The government’s discussions with Meta therefore appear to be aimed at finding a balance between automated moderation and human oversight while ensuring that platforms operating in India remain responsive to the country’s legal and regulatory framework. Further discussions are expected to determine how these safeguards can be implemented without unnecessarily restricting legitimate content or expression.