The Wrong Battle

Pakistan’s information policy has spent years fighting the wrong thing. It has focused on content: identifying hostile material, requesting removal from platforms, restricting VPN access and blocking platforms entirely when the volume becomes unmanageable.

The assumption behind every one of these interventions is that the problem is the post. But it is not; the problem is what the post does to the audience before it is ever taken down, and no takedown request reaches that.

The cognitive mechanism hostile digital narratives operate on is not persuasion. It is habituation. Psychological research on anchoring documents this with consistency: repeated exposure to a frame, even among audiences that actively resist it, gradually normalises that frame as a reference point. The claim need not be believed to do its work.

A citizen who encounters, month after month, a primary online narrative of the state as an engine of administrative failure need not accept each story as true. He needs only to absorb that characterisation as the ambient context from which he evaluates subsequent claims. By the time a specific accusation arrives, the baseline he judges it against has already moved. The state is no longer rebutting a claim. It is trying to restore a prior default that the environment has quietly erased.

What produces that environment is not a mass movement. It is a small and documented infrastructure. Pakistan had nearly 80 million social media users as of early 2026, a figure that grew 25 per cent in a single year. Hostile actors do not need to reach all of them. The operational model relies on clusters of anonymous accounts whose coordinated inauthentic behaviour creates the statistical appearance of organic consensus, diaspora-operated amplifier pages that embed hostile political framing inside community content their audiences already trust, and posts timed to coincide with news cycles when platform algorithms are most likely to surface the content into general feeds.

The effect on the audience is not the product of reach. It comes from repetition and apparent ubiquity. The audience believes it is observing a social reality, not receiving a targeted message.

The platforms are not neutral infrastructure in this. Every major social media platform is optimised at the algorithmic level to surface content that generates engagement. Engagement correlates with emotional arousal. Research by the MIT Media Lab has documented that emotionally charged content spreads approximately six times faster than factually neutral content across social networks. This is not a loophole that hostile actors have discovered. It is the designed behaviour of systems built to maximise time-on-platform, and it produces a structural advantage for content that generates outrage, grief, and fear over content that informs.

A missing person post will outperform a programme delivery update every time the recommendation engine runs. The state is not losing the information contest because its content is weaker. It is competing on a terrain whose architecture was designed by engineers, not regulators, and the architecture does not favour accuracy.

Pakistan’s response to this structural problem has focused almost entirely on the content layer. In February 2024, following a politically volatile election period, the state moved to manage access to X in an attempt to bring a chaotic information environment under control, a measure that effectively remained in place for fifteen months before it was restored in May 2025. Beyond that, Access Now recorded 21 internet disruptions in Pakistan in 2024 alone, the highest figure ever documented for the country, the majority of them area-specific responses to potential disorder, including restrictions imposed during Muharram processions and localised shutdowns in flashpoint areas to prevent riots from escalating.

The Pakistan Telecommunication Authority (PTA) moved to mandate VPN registration for banks, embassies, IT firms and freelancers, and had logged over 30,000 registered VPNs by the time it introduced licensing. Ninety-nine per cent of IT companies reported service disruptions from the regulatory environment this created. When access to X was first conditioned, Surfshark recorded new user acquisition rising by 300-400 per cent in the opening days; Top10VPN documented a 131 per cent spike on the first full day.

The VPN surge is the policy’s own verdict on itself. Platform restrictions do not reduce the audience’s exposure to hostile content. They teach the audience to route around the state. Every major restriction has produced two things the hostile narrative needed: a temporary spike in the visibility of whatever the platform carried, and a durable demonstration that the state’s instinct, when confronted with information it cannot control, is to restrict rather than contest. Restricting access becomes content. It circulates as evidence, on the hostile narrative’s own terms, that the state fears what citizens might see. The cognitive anchor set by that demonstration outlasts any individual piece of removed content.

The regulatory architecture Pakistan has built to manage digital hostile narratives is calibrated to a threat model the technology left behind a decade ago. The threat is not the post. It is the platform’s design, the coordinated infrastructure that exploits it, and the cognitive effect of sustained exposure that neither takedown nor restriction addresses. Regulating which VPNs are licensed does not regulate the algorithm. Removing a piece of coordinated inauthentic content after it has completed its distribution cycle does not lift the ambient frame it has established.

Pakistan has 161 million broadband subscribers and an IT export sector that reached $4.6 billion in FY2026. The country has the technical capacity to engage with the platform layer at a level of sophistication the current policy does not attempt. What the current policy attempts is control of the pipe. The contest is being run inside the pipe, by people who understand its architecture far better than the regulators who built the framework around it.

Technology governance that would address the mechanism requires engaging platforms on their algorithmic design, building independent capacity to detect and attribute coordinated inauthentic behaviour in real time rather than after the distribution cycle has closed, and developing a counter-presence that operates at platform speed with platform-native content.

None of these is communications strategies. They are technical and institutional capabilities Pakistan has not built because the policy framework has treated hostile digital narratives as a content problem that laws and restrictions can manage. The wrong battle, fought with the wrong tools, will keep producing the same result.

Originally published in The News, this article by Amir Jahangir is featured in South Asia Times as an Editor’s Pick.

Amir Jahangir

The writer is a strategic communications expert, media executive, and policy adviser