How Internet Shutdowns Became the New Face of Authoritarianism

During the Salzburg Global Public Policy New Voices Europe cohort 4 meetings and workshops, I gathered with 29 Fellows from 16 countries. Digital disinformation was not merely a talking point in our conversations. It was the lived reality for many people around me. Those conversations deepened a question I had begun exploring in my research at the University of Kassel: What happens when the infrastructure people used to speak, organize, and dissent is not manipulated but simply switched off by the state?  

The justifications governments offer follow a familiar script: national security, public order, stopping misinformation, and protecting elections. Yet Access Now's own research shows that the stated reason rarely aligns with the actual trigger. Shutdowns cluster around protests, elections, and moments when citizens are trying to be heard. That gap between the official explanation and the actual cause is where the real story lies.

The mechanics vary, but the logic holds across different political contexts.

When a government orders a platform to comply or face a ban, and the platform complies, the state learns something useful. It learns that access is leverage. In one documented case, a major social media platform was suspended for seven months after it removed a post by a head of state. The platform returned under conditions requiring local registration, content regulation, and a formal compliance agreement with the government. What began as a shutdown ended as a settlement, and other governments were watching.

In other contexts, shutdown power has been written directly into law, with stated time limits and disclosure requirements. In practice, orders still rely on vague language that falls well short of those standards. One democracy recorded 65 shutdowns in a single year, the highest figure among democratic states globally, despite having a legal framework nominally designed to prevent abuse.

Elsewhere, legislation requires foreign platforms to appoint local representatives and to comply promptly with content removal requests, under threat of heavy fines and service throttling. In a recent case, a platform complied with a court order restricting the account of a major opposition figure during active political protests. The platform did not resist. It complied.

The pattern is consistent regardless of where it appears. Digital access is no longer neutral infrastructure. It has become a pressure point, and states have learned exactly how to use it.

The people who bear the cost of these shutdowns are rarely the ones who ordered them.

When a government shuts down the internet, the immediate casualties are not only political but also economic and social. The global cost of internet shutdowns reached $19.7 billion in 2025, a 156 per cent increase from the previous year. That figure does not capture the consequences for a journalist who cannot file a story, an activist who cannot coordinate a response, or a small-business owner whose income vanishes overnight, with no warning and no recourse.

The United Nations has warned that shutdowns deepen existing inequalities, hitting poorer communities, rural populations, and mobile-dependent users hardest, precisely those with the fewest alternative means of communication and the lowest ability to absorb the disruption.

When shutdowns become routine, something else happens, too. People begin to self-censor. They post less, organize less, and say less, not because the internet is down, but because they have learned it could go down at any moment. That anticipatory silence is part of the cost, and it rarely appears in any economic estimate.

The tools to address this already exist. What is missing is the will to use them.

First, every shutdown order should be publicly disclosed before or immediately upon its issuance. Several countries already have laws requiring this. Where those laws exist but are routinely ignored, the gap between policy and practice is itself the problem.

Second, platforms need publicly stated standards for when they comply with state demands and when they do not. Compliance with every government request is not neutrality. It is a choice with real consequences for the people on those platforms.

Third, the 2011 joint declaration by the United Nations, the Organisation for Security and Co-operation in Europe, the Organisation of American States, and the African Commission on Human and Peoples' Rights established that restricting internet access can never be justified under international human rights law. It deserves more than moral authority. It deserves enforcement.

The internet was built on open infrastructure. The question now is whether those with the power to protect it will act before the pattern becomes permanent.