The world’s largest encyclopedia was overrun by bias and censorship—and pushed me out when I tried to fix it.
Twenty-five years ago, I co-founded Wikipedia, arguably the most important encyclopedia in human history. On Monday, I was indefinitely banned from the site. The story of what happened to me is, in many ways, the story of our censorious times, in which independent thinking is seen as a threat rather than a virtue, and punished as such.
Let me back up.
In early 2000, the internet was a very different place than it is today. It was much freer—tools of censorship weren’t nearly as advanced—but it was also harder to use, and finding information took much longer. The need to make the internet more user-friendly was clear to those of us who spent a lot of time online. We needed a free, fair storehouse of knowledge: an encyclopedia built by, and open to, the public.
It was exhilarating to build Wikipedia at that time. Never before had a global, volunteer-written encyclopedia been shown to work. We applied the principles of open-source software to knowledge: All users would have a seat at the table, everyone would edit each others’ work, and the results would be free for all.
By early 2002, Wikipedia had begun taking off in a big way. Entries already numbered 20,000, and with each new article, Google pushed more traffic our way. Atop each page read the message, “You can edit this page right now!” And people did. Users joined so quickly that the number of Wikipedia articles surpassed 100,000 articles by early 2003.
But as the disastrous effects of the 2000 dot-com bubble spread, the start-up paying me to organize the site ran out of funding. It was just as well. The site was increasingly overrun by trolls, and my co-founder, Jimmy Wales, and I disagreed about this and other problems. While I had started and led the project, he was the CEO. So I left.
Over the following years, I watched in dismay as the site I’d created began to drift from its founding mission. The weakening of Wikipedia’s basic principle—neutrality—was my main concern. Instead of maintaining strict neutrality on the most important and contentious topics, Wikipedia has, over time, become decidedly globalist, academic, secular, and progressive. In other words, it has become the mouthpiece of a relatively narrow liberal establishment, leaving perspectives beyond its borders out in the cold.
For example, rising anti-Zionism in leftist circles means pro-Israel views are disfavored, while pro-Iranian regime sentiments are permitted to spread. Many confessional Christian sources have been disallowed due to the tacit prioritization of secularism in Wikipedia’s principle of neutrality. Hindu organizations have been censored and smeared. And American political topics—about politicians as well as issues like abortion, illegal immigration, and transgenderism—are covered from a broadly progressive-left perspective: The Republican Party, for instance, is classified as being on the “right-wing to far-right” of the political spectrum. And the Democratic Party? “Center to center-left.”
How could this happen to a supposedly neutral encyclopedia that anybody can edit? It goes back to a fundamental problem with the site: Wikipedia has never developed a community charter. Instead, it operates under vague, collaboratively written rules that are interpreted by an all-powerful class of “admin” moderators beholden more to each other than to any constitutional framework. It is ruled by an anonymous mob, and not even a large one; of about 800 administrator accounts, only about 400 are active. Because of that, important contributors have been blocked; pages about socialism and communism whitewashed (or turned into hit jobs, or deleted outright); facts censored in the name of “undue weight” and avoiding “fringe views”; and left-leaning outlets overwhelmingly favored. All this transpires with no mechanisms for real accountability.
Despite this, I never lost my conviction in the platform’s importance. And last year, I saw a major opportunity to propose a reform plan for Wikipedia, the first ever in the project’s history, for the site’s 25th anniversary. I published my Nine Theses on Wikipedia around Reformation Day, in September 2025. The program advocated, among other things, enabling competing articles, abolishing biased source blacklists, and reviving our original policy of neutrality on controversial topics. The theses received a lot of media attention, and I hoped they would spark a movement to reform the platform without requiring my active leadership.
Unfortunately, that did not happen. Two-thirds of the commenters on the proposal were opposed, and among those opposed were some of the most influential voices on the platform, including administrators and experienced editors—the kind of voices that naturally intimidated my supporters.
Instead of maintaining strict neutrality on the most important and contentious topics, Wikipedia has become the mouthpiece of a relatively narrow liberal establishment.
So, this year, I took the lead. I began building a reform group to operate within Wikipedia’s WikiProject ecosystem—a system where teams of contributors can band together to improve the site. Our group was called WikiProject Intellectual Diversity (WPID), and its aim was to guide Wikipedia’s policies toward allowing a more diverse set of views than those currently permitted on the site. Whoever was systematically excluded from Wikipedia, those were the people we focused on: Hindus, confessional Christians, Israelis, American conservatives and libertarians, as well as TERFs, alternative medicine practitioners, and climate skeptics. It was a big tent by design.
I then developed an automated list of the latest policy discussions on the site and began recruiting members to WPID—primarily on X—in compliance with Wikipedia guidelines. Eventually, I had gathered almost 30 volunteers, and we had taken all the steps necessary to be approved as an official WikiProject, including defining the scope of our project and ensuring it wasn’t redundant with any existing ones.
Last Thursday, I submitted our application for approval to the WikiProject Council, a group of volunteers who oversee the ecosystem.
I am no stranger to online controversy, but what happened next surprised me. Last Friday morning, I woke to dozens of comments from users on my WPID application. For simply advocating for a wider use of sources and a more open decision-making process, some critics accused me of trying to grant an “unearned DEI-type lift to far minority, such as far-right, positions.” One editor called WPID “a barely concealed attempt at a coup d’état via canvassing. It’s a sabotage operation.” Another insisted: “Strongest oppose in existence. Far-right extremism is not welcome on Wikipedia. Trans rights are human rights by the way.” Another wrote simply: “Oppose because Larry Sanger.”
I updated my X followers, bemused, telling them about the absurd debate over my project. Then some users noticed this, and accused me of “off-wiki canvassing.” “You are using your platform to influence the vote,” they said. Of course, I never called for participation in the debate; I just said it was happening. Even Jimmy Wales agreed it wasn’t canvassing, writing that my post was “unambiguously fine.”
No matter. On Friday afternoon, I was brought up on charges of canvassing before the Administrators’ Noticeboard—in effect, Wikipedia’s court system. Throughout Saturday and Sunday, I attempted to defend myself, including by explaining that my promotion of WPID on social media had broken no rules and by pointing out that WPID does not aim to disrupt decision-making on Wikipedia but rather to increase participation in it. Not only did this not work, but the extended, rigorous arguments I made were then added to my list of offenses. I was arguing too much! That’s called “bludgeoning,” don’t you know? Apparently, instead of defending myself, I should have issued only mea culpas. Then, one editor pasted the entire thread into Grok—Elon Musk’s AI—and asked for its opinion. Grok’s assessment was more balanced than most of the human comments, noting “pile-on culture” and that my underlying concerns about bias were “not fringe or invented.” Then that answer was posted, circulated, and mocked.
I knew Wikipedia’s disciplinary processes were bad—but I had never experienced them myself. I was tried by a faceless mob. I learned that their greatest anger is reserved for those who refuse to bow in awe of their mighty power.
On Monday morning, while I was on a flight to New York City, I was blocked from the site by a Wikipedia admin who claimed that the mob’s (the “community”) consensus favored my banning. A couple of hours later, I was unblocked by another admin who said a full 72 hours of discussions had to pass before sanctions were implemented.
But the writing was on the wall. By dinnertime, I was blocked again, this time permanently, from the site I had founded, by someone called “ScottishFinnishRadish,” who said of me that “There is general agreement among participants that he has engaged in off-wiki canvassing and is not here to constructively build the encyclopedia.”
Don’t laugh—ScottishFinnishRadish is a very distinguished, honored, and powerful Arbitration Committee member. Wikipedia does not require identity confirmation from its leadership ranks; in fact, because the culture is built to allow anyone who wants to contribute to do so, one could argue it actually discourages self-identification. As a result, only about 15 percent of Wikipedia leadership has disclosed their identities, and many of the rest operate under pseudonyms.
In one of life’s quirky ironies, two hours after I was kicked off the site I founded, I was at a gala dinner in lower Manhattan, accepting a Tablet magazine Sinai Award, which honors individuals who have acted courageously to promote freedom. What was the award for? Co-founding Wikipedia and attempting to reform it.
In my acceptance speech, I told the story I just told you. And as I said then, the last few days have been filled with a mixture of bemusement and shock. But my life has had a lot of controversies like this because, as a general rule, I stand on several principles, no matter who they bother:
- There is an objective truth.
- Knowledge is one of the most important things in life, and it ought to be made available for free, if at all possible.
- Knowledge projects must absolutely be neutral.
I was tried by a faceless mob. I learned that their greatest anger is reserved for those who refuse to bow in awe of their mighty power.
Wikipedia changed the world. But its tragedy is that it grew so large that it sucked most of the air out of competing projects, and even as it rose to dominance, it was taken over by ideologues and shills who co-opted it for propaganda purposes. For those people, the temerity to try to recruit people from outside their weird clique was an unpardonable sin.
Unfortunately for them, I don’t particularly care.
Over the past year, I have fought tirelessly to reform Wikipedia because its founding mission is more important than ever. Information is the most valuable currency in any society, and the ability of citizens to access, evaluate, and learn from a diversity of viewpoints is essential to a free civilization. Yet Wikipedia’s yearslong shift away from that principle—toward ideological gatekeeping and narrative control—undermines the very purpose for which it was created.
So now that I am powerless to fix the platform from the inside, what should I do?
I’ve been weighing lots of ideas. One would be a tool to allow the public to rate and discuss Wikipedia articles. The same system might direct users toward alternatives such as Grokipedia and many other platforms making up the Encyclosphere. In the long term, I’d like to archive all the world’s free encyclopedias, making them available in a single format and sharing them across a truly decentralized network, like the old-fashioned internet. Wikipedia needs a competitor—or, perhaps, a system of competitors.
The company I created may no longer uphold its own founding principles, but people still want and need the kind of knowledge Wikipedia was built to provide. They deserve more than one place to look for it.
