Wikipedia’s AI Problem Could Change the World’s Shared Memory

Wikipedia has become so familiar that it can feel like part of the internet’s furniture. Yet the numbers behind it are extraordinary: 68 million articles have been consulted trillions of times, while Wikipedia is described as the largest knowledge resource ever assembled. Founded by Jimmy Wales and Larry Sanger in 2001, it grew from an unusual experiment into a shared reference point for people around the world.
English Wikipedia alone contains 5.2 billion words. Across more than 300 languages, Wikipedia has 65 million articles. Those figures describe a project built by people who wrote, checked, argued over, and revised information in public. That human effort now faces pressure from the same artificial intelligence systems that depend on vast stores of human knowledge.
From a strange experiment to a global reference
Wikipedia did not begin with the polish that many writers and journalists expected from a major encyclopedia. Richard Cooke wrote that “Most of the people seriously interested in Wikipedia were a handful of academics.” He also said Wikipedia’s “amateurish, homegrown ethos” was off-putting to a lot of writers and journalists.
That roughness was part of the point. While many people in the dotcom boom started online encyclopedias, Cooke observed that “Wales was the only one to hire a philosopher.” The project developed through open participation rather than a single authority deciding what the world should know, and that approach helped create a resource with an enormous reach.
Wikipedia’s importance becomes clearest during moments when people need information fast. The site is accessed with special frequency during natural disasters, when ordinary routines break down and people search for basic facts about places, events, and risks. Its value is not limited to formal research. It also captures the odd mixture of wonder and distraction that defines modern life.
In 2013, an online poster asked what would be difficult to explain to someone from the 1950s. The top answer was: “I possess a device, in my pocket, that is capable of accessing the entirety of information known to man. I use it to look at pictures of cats and get in arguments with strangers.”
The old dream of one world encyclopedia
The idea behind a shared encyclopedia is older than Wikipedia. H. G. Wells wrote essays collected in World Brain that promoted a “Permanent World Encyclopaedia.” But earlier European encyclopedias, including the Great Soviet Encyclopedia and the Enciclopedia Italiana, also showed how knowledge projects could express nationalist and totalitarian intent.
The political stakes were not hidden. The entry on “Fascism” in the Enciclopedia Italiana was credited to Benito Mussolini. George Orwell challenged Wells’ ideas in his 1941 essay “Wells, Hitler, and the World State,” describing Wells as a nostalgic, nineteenth-century liberal rather than a futurist.
The Enlightenment encyclopedia carried its own vision. It was described as a way of seeing and doing, a lens through which a clear, scientific view of all things could come into focus. Alasdair MacIntyre classed “encyclopaedia” as one of the three major approaches of Western ethics. In other words, an encyclopedia is not only a pile of facts. It also reflects beliefs about who gets to record knowledge, how truth is organized, and whose experience counts.
AI puts the human foundation under pressure
The tension between artificial intelligence and Wikipedia has been building for years. In 2018, Google was training AI to write “Wikipedia-like articles.” In 2019, Katherine Maher wrote a New York Times op-ed titled “Without Humans, A.I. Can Wreak Havoc.” The warning now feels tied directly to Wikipedia’s future.
By October 2024, about 5 percent of nearly 3,000 newly created articles on English Wikipedia in a single month contained a substantial amount of AI-generated content. In March 2026, AI content was found to violate Wikipedia’s core principles so often that writing with large language models was banned outright.
The concern is not only whether a machine can produce readable paragraphs. Wikipedia depends on writers who bring judgment, curiosity, and the willingness to challenge errors. The greatest threat to the platform is considered the yearly “whittling down” of its pool of talented writers. If fewer people create and review articles, the encyclopedia loses the human process that gives it value.
A viral X post captured the frustration in one sentence: “I want AI to do my laundry and dishes so that I can do art and writing, not for AI to do my art and writing so that I can do my laundry and dishes.” The complaint points to a basic choice. AI can remove tedious work, or it can replace the creative and careful work that makes shared knowledge useful.
Wikipedia’s blind spots matter too
Even without AI, Wikipedia has never been a perfect record of the world. Just 19% of its biographies are of women. Sandi Toksvig described the site in blunt terms: “Wikipedia is patronising and deeply biased.” That criticism raises a difficult question: what happens when AI learns from a knowledge base that already leaves many people and subjects underrepresented?
Google’s AI Overviews have become as central to the search experience as Wikipedia was until recently. That shift changes how people encounter information. Instead of opening an article, checking its history, and seeing the debate around it, users may receive a summary generated from material they never examine directly.
Wikipedia’s future therefore depends on more than blocking machine-written articles. It depends on keeping people involved, expanding whose lives and work appear in the encyclopedia, and protecting the open process that made the project matter. Its history shows that every grand encyclopedia carries a view of the world. The question now is whether that view will remain shaped by people, or be thinned out by systems trained on what people have already made.
Based on
- Wikipedia often gets taken for granted – but its history is fascinating, and its future is fragile — theguardian.com
- Wikipedians Believed AI Would Come and Go. They Were Wrong | WIRED — wired.com
- Is Wikipedia ‘deeply biased’? Sandi Toksvig says it is – so she’s backing a rival | Wikipedia | The Guardian — theguardian.com




