Zuckerberg Apologises, But India Strikes at Meta’s Legal Core Over Algorithmic Power
In a sharp escalation of its long-running scrutiny of Big Tech, the Indian government on Wednesday informed Meta that the company does not qualify as a neutral “intermediary” under Indian law. The reason, officials said, is straightforward: Meta’s algorithms actively decide which users see which content. The declaration came during high-level meetings in New Delhi even as Meta CEO Mark Zuckerberg conveyed apologies for the presence of child sexual abuse material, deepfake content, and broader operational failures on Facebook, Instagram and WhatsApp.
The development marks one of the strongest challenges yet to the legal protections that have allowed global platforms to operate with limited liability for user-generated content in India. It also transforms a controversy that began with the temporary removal of a single video by Prime Minister Narendra Modi into a deeper debate over algorithmic power, platform accountability and the future of safe-harbour protections.
The Spark: A Viral Video That Vanished
The sequence of events began on 23 July 2026. Amid nationwide student protests over examination paper leaks, Prime Minister Narendra Modi posted a direct, selfie-style video aimed at young people. First uploaded on Instagram and later shared on Facebook, the message sought to reassure students that the government would take stronger action against those responsible for the scandals. The video quickly became one of the most-watched political posts in recent years, drawing hundreds of millions of views.
In the early hours of 28 July 2026, the Facebook version of the post disappeared for users in India for several hours. Viewers were shown a notice indicating that access had been restricted following what appeared to be a legal request. Meta restored the content and publicly described the removal as an error caused by a glitch in its automated artificial-intelligence content filters. The company apologised and later told the government it had introduced additional layers of human oversight for posts by the Prime Minister and other prominent figures.
Indian authorities were unconvinced. The Ministry of Electronics and Information Technology summoned senior Meta executives and described the technical-glitch explanation as inadequate. Officials made clear that the matter was far from closed. The government continued to press for detailed policy and technical accounts of how the removal occurred and what safeguards would prevent a recurrence.
Parliamentary Pressure and the Three-Day Ultimatum
The issue gained sharper political edge when the Parliamentary Standing Committee on Communications and Information Technology, chaired by BJP MP Nishikant Dubey, intervened. The committee demanded an unqualified personal apology from Mark Zuckerberg within three days. Committee members argued that the temporary disappearance of the Prime Minister’s content was not the act of a neutral intermediary but of a publisher exercising editorial control. Failure to apologise, they warned, could lead to the withdrawal of Meta’s safe-harbour protection under Section 79 of the Information Technology Act.
Safe harbour has been the cornerstone of how large social media platforms operate in India and many other jurisdictions. It provides conditional immunity from liability for content posted by users, provided the platforms follow due-diligence requirements, including the timely removal of unlawful material. Without that protection, platforms and potentially their executives become directly exposed to civil and criminal complaints arising from content on their services.
Wednesday’s Meetings and Meta’s Admissions
On 5 August, a Meta delegation led by Global Affairs head Joel Kaplan met IT Secretary S. Krishnan and later IT Minister Ashwini Vaishnaw. According to government sources, Zuckerberg’s team conveyed apologies covering child sexual abuse material, deepfake content and operational errors. Meta also acknowledged that substantial sums had been spent boosting certain types of content and expressed regret for the lapses.
Indian officials responded with a more fundamental position. Because Meta’s recommendation systems select, rank and determine which users receive particular posts, videos and advertisements, the company cannot claim the status of a passive intermediary. In the government’s assessment, active algorithmic curation places Meta outside the traditional safe-harbour framework. Sources indicated that the safe-harbour protections under the IT Act would therefore not apply in the same way.
Why the Distinction Matters
The distinction between a passive host and an active curator goes to the heart of modern platform regulation. Early internet law treated platforms largely as conduits, similar to telephone networks, that merely carried third-party speech. Recommendation algorithms, personalised feeds and paid amplification have fundamentally altered that model. Platforms no longer simply host content; they decide its visibility, reach and commercial value.
By asserting that Meta’s systems take it beyond intermediary status, Indian officials are signalling that the legal protections designed for an earlier era of the internet may no longer fit platforms whose core business depends on shaping what users see. The move aligns with a broader global conversation about whether large social media companies should continue to enjoy the same liability shields when their algorithms play such a decisive role.
Consequences and the Road Ahead
Loss of intermediary status would carry significant practical consequences. Meta could face greater legal exposure for illegal or harmful content, including the possibility of direct complaints and cases against the company or its executives. Compliance costs would rise. The platform’s ability to rely on automated systems without heightened human oversight could come under pressure. More broadly, the position taken by the government could influence how other large platforms operating in India are treated.
Further meetings between Meta and Indian authorities are expected. The parliamentary committee’s three-day window for a personal apology from Zuckerberg remains in play. Officials have also indicated that issues of child safety, paid promotion of problematic material and algorithmic accountability will continue to feature in discussions.
For Meta, India represents one of its most important growth markets. For the Indian government, the episode is part of a longer effort to ensure that global platforms operating at massive scale inside the country accept greater responsibility for the content they amplify and the systems they deploy. The coming days will determine whether the current confrontation produces clearer rules of engagement or opens a more prolonged period of regulatory friction.
The temporary removal of a single video has now become a test case for a much larger question: in an age of powerful recommendation algorithms, what does it mean to be a neutral intermediary, and who should bear responsibility when those systems fail or cause harm?
