In Jantar Mantar protests, a model of all that’s wrong with India’s internet governance

Let’s look at some of these measures and the underlying processes. First, Telegram, the popular messaging platform with 150 million users, was temporarily banned and ordered to disable its message editing feature. The relevant legal provision in the Information Technology Act empowers the government to block public access to “information” on certain grounds, including public order. The court read “information” to be wide enough to include an entire platform instead of specific pieces of content. Analysing the blocking against the Supreme Court’s framing of the proportionality test in Anuradha Bhasin, the court agreed with the government that blocking was the “least restrictive” measure. Much of the court’s finding in the government’s favour hinged on the temporary nature of the order.

This sets a dangerous precedent. Domestically, it normalises the blocking of an entire platform without regard to the disproportionate impact on millions of users, or required scrutiny into whether it would be effective in achieving the desired outcome. The ineffectiveness is made apparent by the fact that people quickly turned to VPNs and moved to other apps following the ban. It’s a classic (and almost literal) example of shooting the messenger. It also signals that there may be a judicial proclivity to uphold it if the banning order is cosmetically compliant and uses the required legal terms such as “temporary” and “narrow tailoring”, even if without meaningful substantiation.

Internationally and economically, we risk being perceived as an unpredictable democracy that may ban a platform as a whole in response to a small fraction of its users misusing it, and interfere at the product-design level, which is bound to negatively impact innovation, security development and ease of doing business.

Next, amid protests, mobile internet services around Jantar Mantar were suspended. When people turned to offline messaging services like BitChat, which work on Bluetooth mesh networks and require physical proximity, the government ordered GitHub to remove source code repositories for BitChat. Globally, research shows that internet access is vital for fundamental rights, particularly during protests, as it aids free expression, access to accurate information, communication with family, and documentation for accountability. Correspondingly, there is no evidence proving its efficacy for public order. Why then should shutdowns have a place in a leading democracy like India? BitChat was blocked merely because it is capable of being misused. If this sliver of possibility was legally sufficient, we would have no AI assistants, SIM cards, email or even Microsoft Office. It is also ineffective and counter-productive. Removing the repositories does not prevent already downloaded apps from working, and only creates security risks by preventing scrutiny of the open-source application. It undermines India’s consistent framing of open-source as a strategic pillar of its approach to AI, contravenes global trends of bolstering decentralised technology as a way to counter the centralised power of a handful of companies, and signals regulatory uncertainty hurting innovation and investment.

On the ground, protestors were being surveilled through AI-powered facial recognition tech (FRT), smart glasses and the “Ikshana” mobile surveillance vehicle, inside which faces of protestors were being matched against a police database. By the police’s own admission in response to Internet Freedom Foundation’s RTI, no rules apply to FRT use, no privacy impact assessment is conducted, and an 80 per cent match is considered positive, which is deeply problematic given the widely documented inaccuracies of FRT. Our Digital Personal Data Protection Act has been enacted but is not fully in force, rendering the constitutionally guaranteed right to privacy without teeth in such high-pressure situations. To grasp how rights-damaging this is, consider the fact that the European AI Act in fact prohibits police from using live FRT in public spaces.

Finally, on content moderation, Parliament threatened Meta with loss of safe harbour — which is the protection online platforms have from liability for third-party content — after the PM’s video became temporarily unavailable. Political chess aside, efforts to rein in Big Tech’s opaque content moderation practices are necessary. But the safe harbour provision is what enables the democratic aspects of the internet to thrive because platforms do not have an incentive to over-censor to save themselves. It is time to look beyond this provision as a tool for platform accountability. To begin with, there should be a focus on establishing stronger systems for algorithmic transparency to scrutinise how our attention is monetised, risk assessments to determine mitigation strategies, and mandatory security features that strengthen user choice and autonomy.

The NEET protests should catalyse the process of making amend(ment)s to India’s digital governance framework. The Information Technology Act has long been in need of an overhaul, internet shutdowns need to become a thing of the past, and the Digital Personal Data Protection Act must be enforced in a consultation-driven manner.

The writer is Asia Pacific Policy Manager and Global Encryption Policy Lead at Access Now