India's safe-harbour threat is more political than legal. That is the real danger

SOME STATEMENTS are more performative than they sound. On August 7th an Indian parliamentary committee gave Mark Zuckerberg, the chief executive of Meta Platforms, three days to issue an unqualified apology for the brief restriction of a video posted by the country's prime minister. The alternative, the committee warned, would be the revocation of Meta's "safe harbour" protection—the legal shield that exempts social-media platforms from liability for the posts of their users. A stern threat is one thing. A workable legal mechanism is another.

The incident that triggered it was small. On July 23rd Meta temporarily restricted a Facebook post by Narendra Modi, the prime minister, addressing student protests over leaked examination papers in the NEET test, one of India's most competitive university entrance exams. The initial restriction page suggested the block came at the request of a third party. Meta later attributed the action to an operational error. The post was restored within hours.

The context was far from trivial. India was in the throes of its largest wave of youth-led protests in more than a decade. The NEET paper-leak scandal led to the resignation of the education minister, Mr Pradhan. The prime minister himself faced ridicule and jokes on social media. Mr Modi's use of Instagram Reels to engage directly with young protesters made the platforms a de facto public square. To restrict a head of state's message in the middle of a civil disturbance is to risk looking like a censor, even if the restriction was a glitch.

The committee's warning about safe harbour is where the story becomes more political than legal. Under Section 79 of India's Information Technology Act of 2000, intermediaries enjoy conditional immunity from liability for third-party content, provided they comply with prescribed due diligence. The Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Rules of 2021 tightened those conditions. Platforms exceeding a user threshold must appoint grievance officers, publish quarterly compliance reports, and remove content within a stipulated time once it is flagged by the state or a court.

Loss of safe-harbour protection does not come from a parliamentary committee's demand for an apology. It is a content-specific affirmative defence, determined by courts. The government can establish a pattern of non-compliance, which may then defeat a Section 79 defence in individual cases. Amending the broader framework to strip an entire platform of safe-harbour rights would require legislative changes. The committee can recommend; it cannot decree. That distinction matters.

To be sure, the legal scaffolding is real. In 2025 the Karnataka High Court dismissed X (formerly Twitter) over a challenge to Section 79 notices. Meta admitted in 2026 to receiving significant payments for boosting certain content to specific audiences in India, a detail that has fuelled accusations that the platform cannot be relied upon as a neutral public square. And the committee's warning came alongside separate scrutiny over Instagram advertisements that Indian regulators allege promoted child sexual-abuse material. The Modi episode may be the spark, but the underlying grievance is structural: a government that views platform governance as an extension of digital sovereignty.

What is Meta's exposure? Facebook, Instagram and WhatsApp together reach more than 1 billion monthly active users in India, the country's largest social-media audience. For the fiscal year ended March 2025 Meta India reported gross advertising revenue of roughly ₹29,400 crore, with net profits of ₹648 crore. India accounts for approximately 2% of Meta's total trailing-twelve-month revenue of $179 billion. The number is big enough to matter and small enough to survive a setback. Even in the unlikely event of a full safe-harbour withdrawal, the company's revenue elsewhere provides buffer. The more immediate risk is political: a loss of regulatory goodwill could translate into slower grievance processing, tighter content rules and a higher cost of operating in a market where advertising rates are already cheap.

The deeper problem is that India is testing a model of platform regulation that other governments will watch. The country's approach does not prohibit speech outright; it makes foreign platforms accountable to Indian law, Indian courts and Indian political preferences. That is a defensible position. States have a legitimate interest in the content that flows through their digital public squares. The trouble is that the same framework can be weaponised against speech that a government dislikes. When Meta first indicated that the Modi post was restricted at the request of a third party, the accusation was almost inevitable: that the government had asked for the takedown. Meta's subsequent explanation, that the restriction was an error, was both plausible and unconvincing.

The incentives on both sides are transparent. The government wants to demonstrate that foreign technology companies do not operate above Indian law. The parliamentary committee, dominated by the ruling BJP, has an electoral incentive to project strength during a period of mass student unrest. Meta wants to avoid a precedent that would expose it to unlimited user-generated-content liability, which would make operation in a country of 1.4 billion people virtually impossible. It apologised. Joel Kaplan, the chief global affairs officer, did so on August 5th. Mr Zuckerberg followed up during a separate meeting, addressing child safety and deepfake concerns. The company's global team remained in India for further talks with the ministry.

That apology may have done more damage to Meta's negotiating position than the original error. The Indian government has learned that a high-profile incident combined with public posturing extracts concessions. Next time it may not need to threaten safe harbour at all.

For investors, the India episode is unlikely to move Meta's stock by itself. AInvest's aggregate signal labels the stock a buy, buoyed by the company's scale and its advertising dominance. Meta's revenue is growing at a clip that the market rewards: in the fiscal second quarter of 2026 revenue reached $60.8 billion, up 22% on the year. India at 2% of revenue is not a valuation breaker. The relevant risk is not a single market but a model. If India's approach to platform accountability spreads to other large democracies, the compliance cost of running a global social network will rise. That cost is real but manageable, so long as the model does not tip towards arbitrary enforcement.

The better question is whether the safe-harbour framework itself is fit for purpose. The 2021 rules, and draft changes that propose reducing the deadline for removing unlawful AI-generated content from 36 hours to three, are moving in a direction that would require platforms to pre-judge content they do not create. That is a recipe for over-removal, the very thing that triggered this crisis. India's stated goal, to protect users from harmful content, will be better served by rules that distinguish between negligence and mistake, between systemic failure and one-off error. A platform that misclassifies a prime minister's video is not the same as one that allows child-abuse material to proliferate. Treating them identically rewards neither users nor the state.

The parliamentary committee wanted an apology. Meta gave one. The real test, which neither side has acknowledged, is whether the rules governing the digital public square can survive the next political cycle without becoming an instrument of it.