Wikipedians Believed AI Would Come and Go. They Were Wrong

Long before the storm of AI struck Wikipedia, the Wikimedia Foundation was making preparations. Wikipedians battened down the hatches and made confident-sounding predictions that it would break and wash over them. Large language models might even improve Wikipedia, it was said, automating more of the housekeeping—formatting, sources, citations, missing links—and so freeing up humans to do creative work. It was not so simple. In 2019, then-Wikimedia Foundation head Katherine Maher wrote an op-ed in The New York Times under the headline “Without Humans, A.I. Can Wreak Havoc.” “Too often,” she wrote, “artificial intelligence is presented as an all-powerful solution to our problems, a scalable replacement for people.” The issue confronting Wikipedia, retaining human control over machines, was the same one soon being faced everywhere.

Already in 2018 Google was training AI to write “Wikipedia-like articles,” and by October 2024, a Princeton University study found that about 5 percent of nearly 3,000 newly created articles on English Wikipedia in a single month (August 2024) contained a substantial amount of AI-generated content. There were signs. ChatGPT was fond of em dashes—parenthetical statements that look like this one—and curly quotation marks. Sometimes, residues from the prompts used to compose articles were left behind (“Here is your Wikipedia article,” they might say.)

At first, Wikipedians tried to shape and control AI, to make its output “acceptable and constructive,” or to fix or remove if it proved otherwise. Generic fluff like “praising a village for its fertile farmlands, livestock, and scenic countryside” and fake sources were weeded out. By March 2026, AI content was found to violate its core principles so often that writing with large language models was banned outright, and their role restricted to translation and minor proofreading.