What Attacks on Wikipedia Reveal about Free Expression

You can learn a lot about a state's view of freedom of expression by how it treats Wikipedia. When states intervene to censor or ban the site, it usually indicates that a significant erosion of rights has already taken place. Systematically tracking how states treat Wikipedia may provide a useful signal for those trying to understand or counter democratic backsliding and rising authoritarianism.

The Wikipedia way, and why it matters

Twenty-four years ago, Wikipedia captured lightning in a bottle. Launched in 2001 as a drafting space for a more traditional encyclopedia, it quickly exploded with activity and left its parent project behind. A few years later, it was one of the most popular websites in the world. Its success inspired a generation of technology writers to speculate about how "the Wikipedia way" could be applied to other domains, proposing models and quotable buzzwords to describe large projects built upon lofty ideals and volunteer labor for a common good. But its success could not be replicated. It stands as an exception, not a formula.

An encyclopedia that anyone can edit, given away for free, written, edited, and run by a large group of volunteers simply should not work, and yet somehow it does. Upon first learning about the "anyone can edit" credo, most people ask the same question: Why is this not garbage? The answer has three parts. First, and most importantly, articles do not come from the minds of contributors; rather, volunteer "editors" (or "Wikipedians") summarize existing publications. Debates over articles, all of which are publicly visible, concern not what is Right or True, but how best to summarize the most reliable sources. Second, decisions are made not by democratic vote or fiat but by a sometimes grueling and incremental consensus-building process, weighing arguments according to how well they align with core principles. Third, as long as there are more people who want Wikipedia to succeed than those primarily interested in advancing an agenda, articles get better and better as time goes by and as more people edit them.

Here's another reason Wikipedia is successful: it began in the United States. The US is where the internet was born and where it flourished, in part, because government, universities, businesses, and organizations worked together, with extensive cross-investment and knowledge-sharing. It is also a country that has regarded free speech, free press, free expression, and academic freedom as core values, upheld as constitutional rights. In the internet age, the US acted quickly to set a legal foundation for the growing web ecosystem by enacting Section 230 of the Communications Decency Act, ensuring innovative platforms could host user-generated content without assuming the author’s legal responsibilities, and giving those platforms the space to make decisions about types of content it wanted to include. For most of the history of the internet, the US has been synonymous with the free and open internet, and Wikipedia is one of its most remarkable outputs.

Perhaps that's why it felt like a major development when powerful people in the US government and an organization capable of influencing policy began using their power to attack and threaten Wikipedia, its institutional host, and its contributors.

Wikipedia requires a confluence of freedoms that does not exist everywhere. As the foremost open knowledge platform, it blurs the lines between media, scholarship, public discourse, digital commons, and popular culture. As a result, Wikipedia provides a unique barometer for a state's attitudes towards free expression. Consider the prospect of starting something like Wikipedia in a nation like China or Russia. First, there is no equivalent to Section 230; instead, platforms are legally responsible for what the users say. This would apply not just to Wikipedia articles, but to the discussion pages where contributors debate how best to improve articles. Combine platform liability with broad intolerance for criticism and dissent, and a project like Wikipedia becomes impossible even before the first article is even written, not because Wikipedia is a tool for activism, but because Wikipedia prioritizes sources with a reputation for accuracy, fact-checking, and editorial independence even when those sources publish things that are politically inconvenient. Beyond the challenge of publishing content, finding reliable sources might also be difficult for some subjects because the state exerts influence or direct control over domestic press and routinely censors independent media when it challenges official positions or narratives. Even in the Chinese and Russian language versions of Wikipedia, which are still hosted in the US and edited by an international community, it is a frequent obstacle to locate good sources on subjects related to politics, health statistics, and both historic and current events in the native language.

Wikipedia as a barometer of free expression

Because Wikipedia summarizes existing publications through a transparent process guided by publicly written principles and does not publish novel ideas or opinions, the act of threatening, censoring, or otherwise attacking Wikipedia is a straightforward extension of an attack on press freedom, academic freedom, and free speech. If a leader does not like what scientists, scholars, journalists, and educators have to say, they will not like what Wikipedia has to say, either. But for those who benefit from sowing distrust in institutions, Wikipedia may be a bigger, easier target, at least rhetorically. It is not formally part of the press, but it is an extremely popular source of news; it is not a university, but it's the most used educational resource in the world. It isn't even a formal organization. There is an organization, the Wikimedia Foundation (WMF), which maintains its servers, develops its software, conducts outreach with the public, coordinates programmatic activities, and manages fundraising, but the foundation does not own or control Wikipedia content. If the foundation were to disappear, Wikipedia would eventually survive, but there would likely be outages, and its infrastructure and operations would suffer from instability. In other words, the foundation is a proxy target.

A challenge for authoritarians in places where the population has already been exposed to Wikipedia (as opposed to, for example, North Korea, which has never permitted access) is that it's immensely popular. Even its harshest detractors will usually confess to occasional use. It is an invaluable resource for information and education, given away for free. It does not display ads, does not accept money for influence, doesn't display sponsored content, and operates almost entirely through donations. In addition to being free to access, its content is freely licensed, usable by anyone else for any purpose. Educators and non-profits have copied Wikipedia, or parts of it, to discs or thumb drives for distribution in regions without internet access. Independent websites, newsletters, and small businesses rely on photos shared by Wikipedia volunteers. It won't sell your data or spy on your web traffic, and it doesn't push you towards material to profit from maximizing your attention. In an age when large technology companies have lost much of their public goodwill, Wikipedia is a striking exception.

Because Wikipedia does not publish original ideas and is so widely liked, attempts to threaten or censor it are rarely a first sign of attacks on free expression, but an indication that an erosion of rights is already taking place. Leaders who prioritize their own interests over the education of their citizens — or worse, fear an educated populace — do not typically begin with such popular not-for-profit resources.

For example, attacks on Wikipedia in Belarus came after a nearly thirty-year decline in free expression. President Lukashenko began consolidating power shortly after his 1994 election, with critical independent media coming under fire by the late 1990s. Harassment of journalists escalated to detentions in the 2000s, and the state seized control of broadcast media. Surveillance, raids, and arrests became common through the 2010s, before reaching a breaking point after Lukashenko won a third election, one widely regarded as neither free nor fair. Crackdowns on protests, journalists, and online expression became more frequent and forceful. Even then, Wikipedia was not censored. Finally, in 2022, editors Mark Bernstein and Pavel Pernikaŭ were arrested for edits against the government's interests.

Censorship — whether bans, threats, or chilled speech — should not be confused with critique, which is an integral part of how the project improves. Sometimes the grievance is shared on Wikipedia, on social media, or occasionally in the news. While anyone can fix mistakes, this ability isn't obvious to everyone, and Wikipedians are used to acting upon external feedback.

Once in a while, critique sparks systemic change. In 2005, someone edited the article about journalist John Seigenthaler to suggest he was involved with the assassinations of JFK and RFK. It was a hoax, but it was not discovered for months. The fallout from the incident directly led to policy changes, tightening quality standards for biographies of living people. It was the first of many high-profile embarrassments that produced an awareness that some areas of the project were not living up to its own goals. Like the encyclopedia articles, Wikipedia's policies and procedures are also subject to never-ending negotiations in pursuit of the core values like "neutral point of view" and "no original research." Sometimes, protracted disputes over how to apply these policies take a long time to wind through bureaucratic dispute resolution processes. The most knotted eventually make it to a case before an elected decision-making body that can enact bans and facilitate future moderation in the topic area (this happened recently on the English Wikipedia with a case related to Israel and Palestine). Wikipedia even keeps track of high-profile critiques of the site in articles with titles like "criticism of Wikipedia" and internal documents like "Wikipedia: systemic bias."

At the most basic level, Wikipedia's greatest strength is also its biggest weakness: an article is only as good as the last editor left it. That person might have added a valuable new section, made a grammatical fix, or replaced the whole page with a joke. Thankfully, Wikipedia's editors have become very good at detecting and correcting "vandalism" and developing technical interventions to protect high-visibility pages, but its open editing model still provides an easy and justified point of concern.

Some criticisms are not actionable, however, as they conflict with the way Wikipedia works. If there is a consensus among the best academic journals, newspapers, magazines, and websites that X is true, objections along the lines of "X is false" will likely be met with a response linking to policy pages explaining how Wikipedia works and a statement like "we say X is true based on summaries of sources A, B, and C. Have we misinterpreted those sources, are those sources unreliable, or do you have other good sources that dispute X?" The discussion is redirected from being about truth or falsity towards an evaluation of publications. Here again, objections to Wikipedia are often extensions of grievances about scientific, academic, or journalistic findings that conflict with a preferred narrative. The words of a politician do not negate a dozen peer-reviewed scientific reviews on a scientific topic, for example, and subjects where reliable sources disagree are presented with perspectives weighted according to their prevalence in the literature. When advanced by members of the public, these grievances are, on the whole, productive, even if they don't result in changes to the article. When they come from officials, on the other hand, with a subtext threatening the exercise of power, it suggests broader institutional discomfort with the norms of free expression.

With a few brief exceptions and mistakes (like the accidental blocking of Wikipedia in the UK in 2008), all of the countries that have banned, censored, or pressured Wikipedia are classified as not free or only partly free by international watchdogs. For example, China, Turkey, Russia, Syria, Iran, and Saudi Arabia consistently rank near the bottom in press freedom indices and have extensive records of jailing journalists and shutting down media outlets. In Saudi Arabia, Syria, and Belarus, individual Wikipedia volunteers have been threatened and even imprisoned. In India, the judicial system has been used to try to influence Wikipedia content and punish volunteers. Since 2002, Reporters Without Borders has published a World Press Freedom Index, based on the idea that good information is necessary for democracy. Out of 180 countries, none of these countries ranks higher than 158.

The Wikipedia Liberty Index

Freedom indices are tools that measure and compare rights and civil liberties, like freedom of expression, across countries and over time. They are often based on detailed surveys of experts with rigorous methodology and published annually. They are invaluable tools for tracking trends, setting benchmarks, identifying problems, and holding leaders accountable, but they struggle to respond to real-time developments, and their breadth can flatten a wide range of actions and attitudes into single scores or summaries.

Perhaps we need a Wikipedia Liberty Index. A Wikipedia Liberty Index corresponds well to other freedom indices, but is not redundant to them. Its comparatively limited scope allows it to be focused and fast. By looking for behavioral, and not just institutional signals, it takes into account attitudes and influence rather than just formal actions. By capturing such a range of actions and statements concerning a single subject, it could be used to compare and foreground specific types of suppression. Wikipedia has no consistent editorial voice or political platform — it is reality as captured by others and summarized by volunteers. As such, efforts to censor it reveal a specific form of insecurity about education and information. Finally, because any one language version of Wikipedia is part of an international project, word of interference spreads quickly across its communities. Real or threatened censorship of Wikipedia is thus a lagging indicator of a state's attitudes towards free press, free speech, academic freedom, and free expression in general, and a real-time indicator of active efforts at suppression.

So, how might a Wikipedia Liberty Index be calculated? The first step is to find, taxonomize, and catalog attacks on Wikipedia. Doing so responsibly will require a system that can quickly vet and organize public information and securely receive and process private reports. It will require a structure capable of accounting for a wide range of possible attacks: nation-level blocks or partial blocks, the creation of ideology-based state alternatives to Wikipedia, targeted disinformation campaigns in Wikipedia articles, efforts to undermine Wikipedia-related institutions, and threats or pressure to silence or chill the speech of individuals or groups of Wikipedia volunteers. For example, we might distinguish the domain of action (social, technical, or legal), category of target (Wikipedia itself, an affiliated organization, or individual contributors), duration of effect (short-term vs. long-term), whether attacks take place on or off Wikipedia, scope (Wikipedia as a whole vs. individual articles), and degree of harm (bans and arrests, reputational harm, coercion, or pressure and propaganda).

One of the fundamental principles of Wikipedia is to "assume good faith" — to operate under the assumption that people contribute out of a genuine desire to collaboratively construct the best summary of knowledge possible. The ideal is that when newcomers make mistakes, experienced volunteers should patiently guide them towards constructive participation. Once in a while, however, it becomes clear that someone is not interested in the project's founding principles, motivated only by their own ideology or self-interest. They may try to frame their arguments using the language of the project's founding principles, but their arguments fall apart under the slightest scrutiny. In extreme cases, such a person should not have the power to uphold and defend the project's rules and values, and that power is taken away. When those entrusted with defending shared ideals act instead in service of themselves, the legitimacy of those ideals — and the institutions built on top of them — begins to erode. That is as true for Wikipedia's governance as it is for any governance.

Thankfully, these cases are relatively rare, and the overwhelming majority of people on Wikipedia participate because they believe in its mission and goals. If that balance were to shift significantly, Wikipedia would stop being Wikipedia.

The author has been involved with Wikipedia in various capacities since 2007. In 2021, he received limited compensation from the Wikimedia Foundation for research-related consulting.

This piece is published with a CC BY-SA 4.0 license by agreement with the author. You may share, republish, or adapt it freely, as long as you credit the author, clearly indicate any changes, link to this original article at Tech Policy Press, and distribute any derivative works with the same license.