The Government Toolkit for Silencing Public Dissent

Key Takeaways

• Blocking livestreams, throttling apps, and cutting mobile data during protests are deliberate choices, not technical glitches.
• The tactics have grown more targeted over time: instead of full blackouts, governments increasingly use throttling, geofencing, and platform-specific blocks that are harder to prove and easier to deny.
• The digital crackdown usually outlasts the protest itself, through surveillance, criminal charges, and the chilling effect on anyone who watched what happened.
• Russia, Serbia, Togo, Nepal, and Bangladesh show different points on the same spectrum of governments treating public dissent as a network problem to be managed.

When a crowd fills a street, the first thing people do is pull out their phones. They livestream, text coordinates to friends, post updates, and document what they're seeing. Governments understand this. In recent years, many of them have started treating a public demonstration not just as a physical event to manage, but as a network event to control. The tools they reach for are getting more varied, more precise, and harder to call out.

The Tools for Quietly Disrupting a Crowd

Early government responses to protest-related internet control often involved full network shutdowns. Those still happen, but they carry real costs: they disrupt commerce, attract international condemnation, and are difficult to deny. So the toolkit has gotten more targeted.

Throttling means slowing down specific platforms rather than cutting them off. A livestream buffers endlessly instead of disappearing, which is harder to prove as censorship. Mobile data cuts mean smartphones still work for calls and texts, but video uploads fail at the moment they matter most. 

Platform blocking takes out Instagram while leaving WhatsApp, or disables Signal while Facebook stays up. Geofencing drops connectivity in the specific area around a public square while the rest of the city stays online. Each of these is harder to attribute to intent than a full blackout and easier for a government to explain as a technical issue or a temporary operational measure.

Togo is a recent example worth noting. When anti-government protests broke out in June 2025, authorities throttled internet access from nine in the morning until midnight, every day, for over two months. They also blocked specific platforms: Facebook, Telegram, Signal, and YouTube. The network was technically still on, which made it harder to classify as a shutdown, but the effect for anyone trying to organize, document, or share what was happening was close to the same.

The Pattern Shows up Across Different Governments

This isn't a tactic limited to one region or one kind of government. Nepal, in September 2025, banned 26 social media platforms after the companies failed to comply with new registration requirements. The protests that followed were met with lethal force, with more than 75 people killed before authorities lifted the bans. 

Bangladesh blocked mobile internet for eleven days in 2024 as a student uprising grew, and then the government fell. In Venezuela, authorities blocked large portions of the internet before, during, and after the 2024 presidential election, detaining journalists and disappearing people who criticized the official result online.

Access Now's global tracking documents dozens of protest-related shutdowns every year across dozens of countries. In the majority of those cases, governments didn't publicly acknowledge the disruption. The connectivity simply disappeared during politically inconvenient moments, and any official explanation came later, if at all.

The Crackdown That Outlasts the Protest

What happens after the crowd disperses matters as much as what happens during the demonstration. This is where the digital toolkit extends beyond network disruption into something more durable.

Russia demonstrates the full arc of this. The government has built a legal infrastructure that makes digital dissent itself a criminal act, not just a risk. People have been jailed for posts made on Telegram, with criminal charges for content that criticized military operations or challenged official narratives. The network blocks are one layer. The criminal consequences for what you posted before the block went up are another.

Serbia has invested in the physical infrastructure to match. State access to surveillance camera footage at public spaces, including transport hubs, gives authorities the ability to identify protest participants after the fact. That capacity changes the calculation for anyone considering showing up. The digital record and the physical record reinforce each other, and both outlast the day of the demonstration.

What Happens to the Right to Raise Your Voice

The right to protest includes the ability to communicate while doing it. A demonstration where people cannot livestream what they're witnessing, cannot reach each other in real time, and cannot share footage with people outside the immediate area is a demonstration that has been quietly diminished. Not banned, just made smaller and less visible.

This is also what the steady erosion of online anonymity makes worse. When people know their posts are traceable, that accounts are monitored, that participation in a protest can be documented and stored, the chilling effect doesn't require a shutdown at all. The awareness that dissent leaves a record is often enough to reduce it. That's the point of building the infrastructure in the first place.

What I keep coming back to is how deliberate all of this is. Throttling a livestream at the moment a protest is gaining visibility is not a coincidence. Cutting mobile data in one district while leaving it on in others is not a network fault. These are choices made by people with access to the relevant switches, made at politically convenient moments. The protest might still happen. What gets controlled is how much of it the rest of the world gets to see.