The Centre’s newly notified rules, which compress social media takedown timelines, underscore the delicate tension between two objectives — the imperative to quash harmful online content swiftly and the risk that expanded enforcement powers could be misused to constrain freedom of speech. Government takedown orders must now be complied with within three hours, a sharp reduction from the previous 36-hour window.
This urgency is not without cause. The recent misuse of Grok, the generative AI tool integrated into X, to create sexualised images of women without their consent exposed a deeply troubling faultline in the social media ecosystem. What made that episode especially disturbing was not merely the abuse itself, but the velocity with which it travelled from fringe experimentation to disturbingly normalised content on one of the world’s largest platforms. So, it is difficult to argue against stronger safeguards. Tech platforms cannot assume that safe harbour protections are automatic or permanent. If they benefit from scale, they must shoulder proportionate responsibility.
Yet, the drastic compression of removal timelines raises profound concerns about due process and free expression. While the takedown obligations are meant to apply to content that violates existing laws, incites public disorder and sexual offences, the scope for interpretation remains wide. In practice, there is a legitimate concern that content exposing governance failures could be swept in if authorities characterise it as threatening public order or spreading misinformation. When compliance windows shrink to mere hours, platforms are likely to err on the side of over-removal. The risk of arbitrary enforcement, and the resulting chilling effect, is real. This tension sits uneasily with the Supreme Court’s landmark ruling in Shreya Singhal v. Union of India (2015), which struck down Section 66A of the IT Act for being sweeping in scope. The Court affirmed that restrictions on online speech must be narrowly and clearly defined. Regulatory mechanisms that prioritise speed without adequate safeguards may inadvertently revive the very concerns the Court sought to remedy. At the same time, leaving content moderation entirely to private platforms has proven inadequate. Companies such as Facebook and X (formerly Twitter) have repeatedly failed to act swiftly against toxic material.
India’s regulatory challenge is therefore not whether to act, but how. The country undeniably needs stronger tools to combat AI-driven abuse and digital harm. But these tools must be anchored in clear statutory standards, transparent processes, and independent review. As India enters an increasingly complex digital era, the guiding principle should remain firm: strengthen protections against online abuse without weakening the constitutional right to speak, dissent, and participate freely in public life. Regulators must build a framework that commands trust from users, platforms, and the courts alike.
Published on February 11, 2026