A groundbreaking study analyzing 21 billion measurements reveals a disturbing truth: internet censorship is rising in 103 countries, signaling a global retreat from digital freedom that extends far beyond traditional authoritarian strongholds.
We tend to think of internet censorship as something that happens âover thereââin China behind the Great Firewall, in Russia under Putinâs digital iron fist, or in North Koreaâs hermetically sealed intranet. But new research from the University of Michiganâs Censored Planet project has uncovered a far more unsettling reality: internet censorship is quietly metastasizing across the globe, spreading to nations weâve long considered free and affecting nearly half of all countries worldwide.
The numbers are staggering. After analyzing 21 billion measurements across 221 countries over 20 months, researchers found increased censorship incidents in 103 nations. This isnât just about the usual suspects anymore. This is about a fundamental shift in how governments worldwide view and control the internet.
The Anatomy of Modern Censorship
What makes this wave of censorship particularly insidious is its sophistication and subtlety. Gone are the days when internet censorship meant simply pulling the plug or blocking a few websites. Todayâs digital authoritarianism comes in many flavors, each carefully calibrated to avoid international outcry while achieving maximum control.
The Michigan study reveals that modern censorship operates through multiple vectors simultaneously. Governments are deploying DNS manipulation to make websites unreachable, implementing deep packet inspection to filter content in real-time, using IP blocking to cut off entire services, and employing more subtle forms of manipulation like throttling speeds to make certain platforms unusable without technically blocking them.
Whatâs particularly cleverâand disturbingâabout these techniques is their deniability. When YouTube becomes unwatchably slow or WhatsApp messages mysteriously fail to send during protests, governments can blame technical issues or network congestion. This plausible deniability makes it harder for citizens to recognize censorship and for international observers to call it out.
The Surprising Offenders
Perhaps the most alarming finding from the research is which countries are increasing their censorship. While nations like China, Russia, and Iran predictably top the list, the study found upticks in censorship in countries that pride themselves on democratic values and internet freedom.
European nations, grappling with concerns about disinformation and online harms, are implementing content filtering systems that would have been unthinkable a decade ago. The UKâs Online Safety Act mandates age verification and content filtering. Germanyâs NetzDG law requires social media platforms to remove âillegal contentâ within 24 hours or face massive fines. France has passed laws allowing the government to order the removal of content within one hour.
Even more surprising are the findings from countries rarely associated with internet censorship. Norway, often ranked among the worldâs freest countries, showed instances of ISP-level blocking. Iceland, despite its reputation as a digital freedom haven, has implemented filtering systems. These arenât wholesale censorship regimes, but they represent a troubling erosion of the principle that the internet should be free and open.
The Democracy Paradox
One of the studyâs most thought-provoking findings is what researchers call the âdemocracy paradox.â Many democratic nations are implementing censorship measures in the name of protecting democracy itself. Whether itâs combating disinformation, preventing election interference, or protecting children from harmful content, democracies are increasingly willing to compromise on internet freedom for other values.
This creates a troubling precedent. When democracies implement content filtering infrastructureâeven with the best intentionsâthey create tools that can be misused by future governments or in times of crisis. The technical capability to censor, once established, rarely goes away. It only expands.
The European Unionâs proposed Chat Control regulation, which would require scanning of private messages for illegal content, exemplifies this paradox. Proposed as a child protection measure, it would create a surveillance infrastructure that could fundamentally undermine encrypted communications. The road to digital authoritarianism, it seems, can be paved with good intentions.
The Pandemic Acceleration
The COVID-19 pandemic marked a turning point in global internet censorship. Under the guise of combating medical misinformation, governments worldwide expanded their content moderation powers. While some of these measures were arguably necessary for public health, many countries have retained and expanded these powers long after the health emergency passed.
The Michigan data shows a clear acceleration of censorship incidents beginning in 2020. Countries that had never previously engaged in systematic internet filtering suddenly developed the capability and legal framework to do so. Emergency powers became permanent features. Temporary measures became the new normal.
Vietnam used pandemic powers to crack down on political dissent. India expanded its ability to order content takedowns. Brazil created new mechanisms for judicial content removal. What started as public health measures evolved into broader content control systems that persist today.
The Technology Arms Race
The technical sophistication of censorship is evolving rapidly. The Michigan study tracked how censorship techniques are becoming more targeted and harder to detect. Instead of blocking entire platforms, governments are now surgically removing specific pieces of content. Instead of obvious error messages, users encounter mysterious connection timeouts or infinite loading screens.
Artificial intelligence is supercharging these capabilities. Machine learning systems can now identify and block prohibited content in real-time, across multiple languages and formats. Facial recognition can link online accounts to real-world identities. Behavioral analysis can identify potential dissidents before they even post problematic content.
But technology is a double-edged sword. As censorship techniques become more sophisticated, so do circumvention tools. The Tor network, despite repeated attempts to block it, continues to provide anonymous internet access. New VPN protocols emerge faster than governments can block them. Decentralized networks promise censorship-resistant communication.
This creates an exhausting arms race where citizens must constantly adapt to maintain their digital freedoms, and governments pour resources into ever more sophisticated control mechanisms. The casualty in this war is the average internet user, who just wants to browse, communicate, and access information without navigating a complex technical and legal minefield.
The Economic Consequences
The rise in global internet censorship isnât just a human rights issueâitâs an economic one. The Michigan studyâs findings suggest that increased censorship correlates with decreased innovation, reduced foreign investment, and slower economic growth.
When countries restrict internet access, they cut their citizens off from global knowledge networks. Researchers canât access international journals. Entrepreneurs canât use global platforms. Developers canât contribute to open-source projects. The cumulative effect is a kind of digital brain drain, where the most talented and connected individuals either leave or find their potential severely limited.
For multinational companies, navigating this fragmenting internet becomes increasingly expensive and complex. Products must be redesigned for different regulatory environments. Content must be filtered differently in each market. The promise of the internet as a global platform is giving way to a patchwork of national internets, each with its own rules and restrictions.
The Normalization Effect
Perhaps the most dangerous aspect of this global rise in censorship is how quickly it becomes normalized. The Michigan study found that once countries implement censorship infrastructure, they rarely roll it back. Instead, censorship tends to expand over time, covering more content, more platforms, and more users.
Citizens, too, adapt to censorship in troubling ways. The research shows that in countries with persistent censorship, users gradually reduce their online activity, self-censor their communications, and accept restricted access as normal. The vibrant, chaotic, free-wheeling internet of the early 2000s is being replaced by a sanitized, surveilled, and controlled digital environment.
This normalization is particularly pronounced among younger users who have never known an uncensored internet. For them, VPNs, careful keyword selection, and limited platform access are just part of how the internet works. The idea of a free and open internet becomes almost mythicalâsomething that existed in the past but isnât realistic in the âreal world.â
Regional Variations
The Michigan study reveals fascinating regional variations in censorship techniques and targets. In the Middle East, censorship often focuses on religious and political content. In Africa, itâs frequently tied to election periods and political unrest. In Asia, economic and social content faces heavy filtering. In Europe and North America, the focus is on illegal content and disinformation.
These regional differences matter because they show how censorship is adapted to local contexts and justified through local concerns. Thereâs no single model of internet censorshipâinstead, thereâs a diverse ecosystem of control mechanisms, each tailored to specific political, cultural, and social contexts.
Latin America shows particularly interesting patterns, with censorship often spiking during periods of political instability. Countries like Venezuela and Nicaragua have developed sophisticated censorship capabilities, while others like Chile and Uruguay maintain relatively open internets. This diversity within regions shows that censorship is not inevitableâitâs a choice that governments make.
The Path Forward
The Michigan studyâs findings paint a sobering picture of the global internetâs future. With 103 countries showing increased censorship, weâre witnessing not just isolated incidents but a fundamental shift in how governments relate to the internet. The question is no longer whether the internet will remain free and open, but how much freedom we can preserve in an increasingly controlled digital world.
Yet the situation is not hopeless. The same research that documents the rise of censorship also shows pockets of resistance and resilience. Some countries have resisted the temptation to implement censorship infrastructure. Others have rolled back restrictive measures after public pressure. The battle for internet freedom is far from over.
Whatâs needed now is renewed vigilance and commitment to digital rights. This means developing better circumvention tools, creating international norms against censorship, supporting digital rights organizations, and educating users about their digital freedoms. It means pushing back against the normalization of censorship, even when it comes wrapped in legitimate concerns about safety or security.
The Stakes
The University of Michiganâs research is more than just an academic exerciseâitâs a warning. When nearly half the worldâs countries are increasing internet censorship, weâre not just seeing a policy trend. Weâre witnessing a fundamental reshaping of human communication and knowledge sharing.
The internet was supposed to be different. It was supposed to route around censorship, to make information free, to connect humanity in unprecedented ways. That vision is under threat, not from a single authoritarian government but from a global trend toward control and surveillance.
The 21 billion measurements collected by the Michigan team tell a story of gradual erosion rather than dramatic collapse. Internet freedom is not dying in a single catastrophic event but through thousands of small restrictions, each perhaps justified in isolation but collectively representing a fundamental threat to digital rights.
As we process these findings, we must ask ourselves: What kind of internet do we want to leave for the next generation? One where nearly half of all countries engage in systematic censorship? Or one where digital freedom is protected as zealously as any other fundamental right?
The data is clear: internet censorship is rising globally, affecting nearly half of all countries. This isnât a distant threat or a problem for other people in other places. Itâs happening now, everywhere, in ways both obvious and subtle. The University of Michiganâs research provides us with a map of this new reality. What we do with that knowledge will determine whether the internet remains a tool for human liberation or becomes historyâs most sophisticated system of control.