India asks Meta to align policies with Indian laws, fix deepfake gaps
Synopsis
India's government isn't asking Meta to rewire its algorithms — it wants accountability layered on top of them. The push for human review of deepfakes involving verified accounts, combined with a zero-tolerance line on child abuse material, signals that New Delhi is moving from passive oversight to active compliance enforcement on one of the world's largest social platforms.
Key Takeaways
The Indian government has asked Meta to align content policies with Indian laws and cultural context, as of 11 August .
Authorities want a 'human-in-the-loop' review mechanism for suspected deepfake content, replacing sole reliance on automated systems.
Content from prominent verified accounts must not be auto-classified as deepfakes — human review required before action.
Meta has been asked to prevent removed deepfake content from resurfacing via repeat uploads.
The government reiterated zero tolerance on child sexual abuse material ( CSAM ) as a top priority.
Officials have also sought greater algorithm transparency around how content is ranked and recommended to Indian users.
The Indian government has formally asked Meta to align its content policies and enforcement mechanisms with Indian laws and the country's cultural context, according to reports dated 11 August. Officials clarified that the engagement does not seek a wholesale overhaul of the social media giant's algorithms, but rather targeted compliance improvements.
Key Demands on the Table
Discussions between the government and Meta are ongoing, with a sharp focus on three areas: strengthening safeguards against harmful content, improving regulatory compliance, and enhancing transparency around content moderation and recommendation systems. Officials are understood to be seeking greater clarity on how content is surfaced, ranked, and recommended to users across Meta's platforms.
Deepfake Safeguards at the Centre of Talks
A central concern flagged by authorities is the growing threat posed by deepfake content. The government has urged Meta to introduce a 'human-in-the-loop' review mechanism for suspected deepfakes, ensuring that takedown decisions are not driven solely by automated systems. Officials argue that human oversight is essential to prevent erroneous moderation and strengthen accountability.
Notably, the government has also conveyed that content shared by prominent verified accounts should not be automatically classified as deepfakes. Such posts, authorities said, must undergo appropriate human review before any enforcement action is initiated. Additionally, Meta has been asked to implement specific measures to prevent deepfake content — once identified and removed — from resurfacing through repeat uploads.
Zero Tolerance on Child Abuse Material
Authorities reiterated a zero-tolerance stance on child sexual abuse material (CSAM), stressing that there can be no compromise on detecting, removing, and acting against such content. The issue remains a top priority in the government's ongoing engagement with the company.
Algorithm Transparency Sought
Beyond content moderation, officials have sought increased transparency around the systems that influence content visibility and user engagement on Meta's platforms. The government wants to understand how Meta's recommendation algorithms function and how they shape what Indian users see. This comes amid a broader global push for platform accountability, with regulators across the European Union and other jurisdictions also pressing major social media firms for algorithmic disclosure.
The discussions are expected to continue, with both sides working toward a framework that balances regulatory compliance with operational feasibility for the platform.
Point of View
But its emphasis here signals frustration with the pace of enforcement. What is less discussed is the algorithm transparency demand: getting Meta to disclose how it ranks content for Indian users would be a significant regulatory precedent, and one the platform has resisted elsewhere. The government's insistence that it is not seeking an algorithm overhaul may be tactical — framing the ask as transparency, not control, makes it harder for Meta to refuse outright.
NationPress
11 Aug 2026
Frequently Asked Questions
What has the Indian government asked Meta to do?
The government has asked Meta to align its content policies and enforcement mechanisms with Indian laws and cultural context. Key demands include introducing human review for deepfake content, preventing removed deepfakes from resurfacing, maintaining zero tolerance on child sexual abuse material, and providing greater transparency on how its algorithms rank and recommend content to Indian users.
What is the 'human-in-the-loop' mechanism the government wants for deepfakes?
It is a review process where human moderators — not just automated systems — evaluate suspected deepfake content before any takedown action is taken. The government wants this to prevent erroneous removals, particularly for content posted by prominent verified accounts, which should not be auto-classified as deepfakes.
Why is the government concerned about deepfake content on Meta's platforms?
Deepfakes pose a growing threat to public discourse, electoral integrity, and individual reputation. Authorities are particularly concerned that purely automated moderation could wrongly flag legitimate content from verified public figures, making human oversight essential for accountability.
What is the government's stance on child sexual abuse material on Meta?
The government has reiterated a zero-tolerance position on child sexual abuse material (CSAM), stressing there can be no compromise on detecting, removing, and acting against such content. It remains the top priority in the government's ongoing engagement with Meta.
Is the government asking Meta to change its recommendation algorithms?
No. Officials have clarified they are not seeking an overhaul of Meta's algorithms. Instead, they are asking for greater transparency around how the systems function — specifically how content is surfaced, ranked, and recommended to users in India.
