Can Meta lose safe harbour over AI moderation? What Sec 79 of the IT Act says

Instagram's temporary removal of a PM Modi video has revived questions over intermediary liability, AI-led content moderation and the limits of safe harbour protection under Section 79 of the IT Act

Instagram, a subsidiary of Meta, recently temporarily removed a video featuring Prime Minister Narendra Modi addressing student protestors over the NEET Exam paper leak. Meta attributed the removal to an automated AI moderation filter, framing the incident as a technical glitch rather than a deliberate move.

 

In response, the Ministry of Electronics and Information Technology (MeitY) convened a Parliamentary Standing Committee chaired by ruling party MP Nishikant Dubey. The central argument advanced by the Standing Committee is that platforms exercising algorithmic control over content cannot logically claim to be neutral or passive channels under the Information Technology Act. Consequently, the Committee argues that such platforms fail to meet the foundational criteria required to claim safe harbor protection under Section 79 of Information Technology Act, 2000.

 

To address the situation, the Parliamentary Committee has put forward two major demands to Meta: A direct, formal apology from Meta CEO Mark Zuckerberg and the revocation of Meta’s safe harbor immunity under Section 79 of the Information Technology Act, 2000.

 

The debate, however, has redirected attention to the larger question of whether the controversy could reshape the interpretation of Section 79 of the Information Technology Act, 2000 (Safe Harbor) and the extent to which Parliament can exert pressure on intermediaries regarding content moderation decisions.

What exactly is 'Safe Harbor' under the IT Act, 2000?

Safe Harbour Protection for Digital Intermediaries: The principle of safe harbour provides that an online platform is not held liable for third-party or user-generated content. Under this framework, the intermediary functions solely as a service provider, offering a platform to users to publish and showcase their content. However, to maintain this protection, the company is required to perform due diligence under Section 79(2)(c) of the Information Technology Act and remove or filter out content that violates guidelines prescribed by the appropriate authority.

 

The Safe Harbour principle does not grant an intermediary absolute immunity; rather, it operates subject to specific legal conditions.

 

Section 79(3) specifies the precise circumstances under which an intermediary’s safe harbour immunity can be revoked: Abetment or Conspiracy: Immunity is waived if the intermediary has conspired, abetted, aided, or induced - whether through threats, promises, or other means - the commission of an unlawful act.

Failure to Take Down Unlawful Content Upon Notice: Immunity is waived if the intermediary, upon receiving actual knowledge or notification from the appropriate government or its agency, fails to expeditiously remove or disable access to the unlawful material on its resource, while ensuring that evidence is preserved.

Consequently, the protections granted to social media companies and digital platforms are not absolute, but are strictly conditional upon adherence to statutory obligations.

How does India's safe harbour framework differ from the US'?

India’s safe harbour framework differs significantly from that of the United States. In the US, Section 230 of the Communications Decency Act (CDA) and Section 512 of the Digital Millennium Copyright Act (DMCA) grant broad, near absolute immunity from civil claims regarding user-generated content. American platforms are explicitly protected even when they actively choose to moderate, remove, or retain content.

 

In contrast, Indian law treats platforms strictly as passive conduits rather than active publishers or distributors. Under Section 79 of the Information Technology (IT) Act, an intermediary must function strictly as a passive host: it must not select the recipient, initiate the transmission, or alter the hosted data. While US jurisprudence has interpreted safe harbor broadly to balance free expression with intermediary liability. Zeran v. America Online, Inc. (1997), established that platforms retain complete legal immunity even when exercising editorial controls like publishing, removing, or retaining user posts. Indian legislation and judicial precedent have progressively narrowed this protection.

What did the Supreme Court say in the Shreya Singhal case?

In Shreya Singhal v. Union of India, the Supreme Court limited the scope of Section 79(3)(b) by holding that an intermediary’s obligation to remove content is triggered only upon receiving a formal court order or a government notification grounded in reasonable restrictions under Article 19(2) of the Constitution. The Apex Court sought to prevent intermediaries from acting as private adjudicators of legality. Exposing platforms to uncertain liability creates an incentive for precautionary takedowns, ultimately chilling lawful expression guaranteed under Article 19(1)(a). Preserving this balance requires clear operational guidelines to prevent arbitrary or discriminatory enforcement.

How has the regulatory framework evolved since Shreya Singhal?

Subsequent executive actions, however, have expanded the operational reach of Section 79(3)(b). In October 2023, the Ministry of Electronics and Information Technology (MeiTY) issued an advisory authorising various ministries, state governments, and police authorities to issue blocking directives under Section 79(3)(b).

 

This was further operationalised in October 2024 with the launch of 'Sahyog', an administrative portal enabling authorities to issue and upload blocking orders directly- a mechanism that risks expanding administrative censorship beyond the limits established in Shreya Singhal.

 

Further narrowing safe harbour protection, the Delhi High Court held in Christian Louboutin SAS v. Nakul Bajaj & Ors. that active participation in promoting, curating or structuring third-party content forfeits Section 79 immunity. The Court reaffirmed that safe harbour protection applies exclusively to passive intermediaries operating as mere conduits, establishing specific operational criteria for e-commerce entities seeking statutory protection.

Does AI moderation automatically make a platform liable?

That is now one of the central questions in the debate.

 

Modern social media platforms do not merely display content chronologically. Algorithms decide what users see, recommend posts, rank material and identify content for restriction or removal. And increasingly, a lot of this behind-the-curtains heavy lifting is being done by AI tools.

 

The central legal debate is no longer whether intermediaries should enjoy immunity, but whether broad statutory protection can coexist with heightened state expectations regarding automated content moderation. To create a modern, balanced legal framework, policymakers must align platform accountability and free speech with intermediary protections.

 

The core issue extends beyond individual platforms like Meta; it centers on whether the existing Section 79 architecture can effectively govern modern, algorithm-driven, and AI-managed systems that actively curate and shape public democratic discourse.