The Indian government is reinterpreting Section 79 of the Information Technology Act to argue that Meta’s algorithmic recommendations constitute publishing, a move that could strip the company of safe-harbour immunity. The dispute, reported after meetings on August 8, 2026, follows Meta’s temporary restriction of a video by Prime Minister Narendra Modi, which the company called a technical error.
The government’s legal theory, if adopted, would treat any platform that algorithmically selects content as a publisher, exposing it to liability. Meta has not lost its immunity, but the next round of talks could formalise demands for human review and faster takedowns.
India’s executive branch is testing a legal theory that could rewrite platform liability without a single vote in Parliament. The vehicle is Section 79 of the Information Technology Act, and the trigger was a temporary restriction of a prime minister’s Facebook video. The government now argues that when Meta’s algorithms decide which content reaches which users, the company is no longer a passive host. It is a publisher, and it should lose the legal immunity that intermediaries enjoy. The dispute, which surfaced in meetings on August 8, 2026, is not just about one moderation mistake. It is about whether India can redefine the meaning of an intermediary through enforcement posture, turning algorithm design into a regulatory trigger. The next round of talks, expected within days, could formalise demands that would force Meta to overhaul its content moderation architecture—and set a precedent for every platform that ranks content.
The legal theory that could sweep in every ranked service
The meetings between Meta and the Ministry of Electronics and IT (MeitY) were reported on August 8, 2026, with follow-up discussions expected the following week. The government is assessing whether Meta’s recommendation systems compromise its intermediary status under Section 79 of the IT Act. Officials contend that platforms performing active curation—deciding which content reaches which users through recommendations or paid promotion—assume a publishing function and must accept corresponding legal responsibility.
Apar Gupta built the Internet Freedom Foundation to defend digital rights in India. Now he is warning that the government’s reading of Section 79 could hand it a lever over every ranked service. “Here, a contrary reading as incorrectly argued by the government would strip search ranking, spam filters, app stores and streaming recommendations of immunity, and hands government a lever over every ranked service.” Legal experts agree that courts, not regulators, hold ultimate authority to determine intermediary liability status, and that stripping safe-harbour immunity may require statutory amendment rather than administrative action.
| Entity | Current rule | New rule | Effective date |
|---|---|---|---|
| Social media platforms | Section 79 safe harbour if not initiating, selecting, or modifying content | Government argues algorithmic recommendations constitute selecting content, ending immunity | Under discussion |
| Significant social media intermediaries | IT Rules 2021 require automated tools for CSAM with human oversight | Government wants at least two human reviewers for decisions on prominent public figures | Not yet formalised |
| Platforms hosting synthetic content | February 2026 amendments mandate labelling of deepfakes with provenance metadata | Government views Meta’s labelling as insufficient, demands broader measures | Already in effect; compliance under scrutiny |
| Platforms hosting specified unlawful content | Three-hour takedown timeline under February 2026 amendments | Government enforcing timeline strictly, creating tension with human-review demands | Already in effect |
| Source: Indian Information Technology Act, 2000; IT Rules, 2021 and February 2026 amendments; government statements | |||
The government’s position rests on a distinction between passive hosting and active curation. Officials argue that when a platform’s systems decide which content is shown to which users—through ranking, recommendations, or paid promotion—the platform is making editorial choices. Those choices, the government contends, move the company beyond the role of an intermediary into the role of a publisher, with corresponding legal liability for the content it amplifies.
The closest Western analogue is the Section 230 debate in the United States, but India’s dispute is being driven through administrative interpretation rather than statutory reform. That makes it a faster-moving model for feed-ranking liability than most U.S. rules allow. The government is pressing for human reviewers fluent in Indian languages and familiar with local political dynamics, especially for cases involving prominent public figures. According to disclosures cited by Gupta, 87% of Meta’s misinformation-classification budget went to English-language work, leaving significant gaps in Indian-language content moderation.
Jyoti Panday, an independent tech-policy researcher, argues that recommendation engines can push platforms beyond passive hosting into publisher-like responsibility when they actively amplify unlawful material. Joel Kaplan, Meta’s chief global affairs officer, apologised to the Indian minister for the Modi-video restriction, signalling direct executive-level engagement with MeitY. The practical exposure for Western users is indirect but real: Meta’s India-specific moderation changes can affect how the company tunes detection and ranking globally. The immediate risk is inconsistent removal or overblocking in mixed-language content.
The legal theory is clear. What remains unresolved is whether the government can enforce it without a court ruling—and what that means for every platform that ranks content.
A regulatory gap that enforcement is filling
India lacks a modern platform-accountability law of the kind the EU built with the Digital Services Act. The IT Act was written in 2000, when intermediaries were passive conduits. The government is now stretching that old framework to address algorithmic amplification, synthetic media, and language-coverage gaps—using the only tool it has: the threat of revoking safe harbour.
This is not a Meta-specific dispute. Officials have indicated they may engage other platforms in a broader review of intermediary definitions and compliance obligations. The mechanism is platform ranking itself. If India treats algorithmic selection as publishing, every major feed-based service operating there inherits a new legal risk calculus around recommendations, paid promotion, and automated moderation.
The next government-Meta meeting, expected within days, is the first moment the pattern could break. Or confirm itself for another cycle.
Beyond the headline
The Power Behind It
India’s executive branch is using enforcement posture to redefine what an intermediary is, without waiting for Parliament or a court ruling. The government is telling platforms that algorithm design itself can trigger publisher liability, turning a technical architecture question into a regulatory weapon.
The Timing
The Modi-post incident gave officials a concrete, politically legible example to attach to deeper concerns about deepfakes, labels, and language coverage. That makes the current window different from earlier policy debates: an abstract legal theory now has a visible trigger inside the government.
The Reach
Meta is the named target, but the mechanism is platform ranking itself. If algorithmic selection is treated as publishing, every feed-based service—from search to short-video apps—faces a new legal risk calculus around recommendations, paid promotion, and automated moderation.
What India’s reinterpretation means for platforms, investors, and users
With the next round of talks imminent, Western companies, investors, and advocates face three distinct decisions.
- Western tech company operating in India
You must re-evaluate content moderation policies, algorithmic design, and legal compliance strategies. The government’s interpretation could expose your recommendation systems to publisher liability. Begin mapping which features algorithmically select or promote content, and prepare for demands for human review and faster takedowns. Consult local counsel on whether your current due-diligence practices satisfy Section 79 conditions under the new reading.
- Global social media platform executive
Monitor India’s regulatory developments closely. If the administrative reinterpretation sticks, it could become a model for other jurisdictions in the Global South. Assess whether your platform’s feed-ranking, recommendation, and moderation systems could be similarly challenged elsewhere. Prepare for possible global adjustments to platform design and content governance.
- US-based investor with APAC emerging market exposure
Increased regulatory scrutiny and potential liability for major tech platforms in India could impact valuation, operational costs, and market stability. Review portfolio companies with significant Indian user bases and algorithmic content delivery. The risk is not just Meta—any platform that ranks content could face similar pressure, raising compliance costs and legal uncertainty.
- Digital rights advocate focused on content moderation
The demands for human oversight and faster takedowns, combined with the reinterpretation of intermediary liability, could significantly impact free speech and due process. Analyze the implications for user rights, advocate for balanced policies that protect expression, and monitor for over-blocking or censorship, especially of political satire and protected speech.
FAQ
Can Meta actually lose safe harbour immediately?
Not without a court finding or legislative amendment. The existing legal structure ties intermediary protection to statutory conditions, but a blanket executive removal of Section 79 immunity is not clearly established. Legal experts note that courts, not regulators, hold ultimate authority to determine intermediary status, and stripping safe harbour may require statutory change rather than administrative action.
What would change for ordinary users if India tightens ranking rules?
The most likely user-facing changes are slower content review, more automated label prompts, more frequent appeals, and potential over-removal of borderline political, satirical, or AI-generated posts. These effects come from platform compliance design, not from a direct user-law change. Users may experience inconsistent enforcement, especially in Indian languages where moderation tools are weaker.
Which other platforms could be pulled into the review?
The government has indicated it may engage other platforms after reviewing Meta’s discussions, but no firm list or timetable exists. If MeitY formalises the review, search engines, app stores, short-video services, and any recommendation-heavy product could face the same legal theory. The mechanism is algorithmic selection itself, not any one company’s conduct.
Explainer
- Section 79
- Section 79 of India’s Information Technology Act, 2000, grants intermediaries safe-harbour protection from liability for third-party content. The protection applies only if the intermediary does not initiate, select, or modify the transmission and complies with due-diligence guidelines. The current dispute centres on whether algorithmic recommendations constitute “selecting” content, which would revoke immunity.
- Intermediary liability
- Intermediary liability refers to the legal responsibility of platforms that host or transmit user-generated content. In India, the framework is built on the idea that passive conduits should not be liable for what users post, but active curation can change that status. The government’s reinterpretation tests whether algorithmic ranking moves a platform from passive host to publisher.
- Algorithmic recommendations
- Algorithmic recommendations are automated systems that decide which content to show to which users, based on signals like engagement history and content characteristics. The Indian government argues that this selection process is a form of publishing, not neutral hosting. If that view prevails, any platform that uses recommendation algorithms could lose intermediary immunity.
- Deepfakes
- Deepfakes are synthetic media created or altered using artificial intelligence, often to depict events that never happened. India’s February 2026 IT Rules amendments require platforms to label such content prominently and include provenance metadata. The government views Meta’s current labelling as insufficient and is pushing for broader safeguards, especially for content involving public figures.
- IT Rules
- The Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Rules, 2021, impose specific obligations on significant social media intermediaries, including appointing compliance officers and using automated tools for certain illegal content. The February 2026 amendments added faster takedown timelines and synthetic-media labelling requirements, which are now at the centre of the government’s enforcement push.
RELATED STORIES
