Wikipedia banned AI-written articles in March 2026 as the site lost almost a quarter of its readership, Wired reports

What happened

Wikipedia's volunteer editors first tried to manage artificial intelligence rather than shut it out, according to a book excerpt published by Wired. By March 2026, AI content was found to break the encyclopedia's core principles so often that writing with large language models was banned outright. AI tools are now limited to translation and minor proofreading.

The excerpt places the ban alongside a steep drop in Wikipedia's readership. It suggests many readers may now be turning to AI summaries instead. For more coverage of how these tools are changing everyday technology, see our AI and tech section.

How AI showed up on Wikipedia

An early study of new articles

Wired reported that Google was training AI to write "Wikipedia-like articles" as early as 2018. In October 2024, a Princeton University study looked at nearly 3,000 new articles created on English Wikipedia in August 2024. It found that about 5 percent of them contained a substantial amount of AI-generated content.

Telltale signs

According to the excerpt, editors noticed patterns. ChatGPT tended to use em dashes and curly quotation marks. Some articles still contained leftover text from the prompts used to create them, such as a line telling the reader, "Here is your Wikipedia article."

From cleanup to a ban

At first, editors tried to make AI output acceptable or remove it when it was not. Generic filler, such as text praising a village for its farmland and scenery, was weeded out, along with fake sources. That effort ended with the March 2026 ban, Wired reported.

A shrinking audience

Pew's anniversary numbers

Wikipedia marked its 25th anniversary in January 2026. Pew Research measured the site's reach for the occasion and found that Wikipedia was viewed 1.9 trillion times between 2015 and 2025. That works out to an average of 508 million views per day, according to the excerpt.

A 'wake-up call' from a longtime editor

Shortly before the anniversary, longtime Wikipedian Christophe Henner published a detailed "wake-up call" for the Wikipedia movement. He wrote that since 2022 the encyclopedia had lost almost a quarter of its readership, which is more than 1.1 billion visits per month. The excerpt says those missing readers were presumably looking for AI summaries instead.

Wired also noted that Google's AI Overviews quickly became central to search. Wikipedia still matters as a source behind those answers, but readers now reach it one step removed.

Why editors are worried

A much-read essay titled "Death of Wikipedia" warned about a possible downward spiral. If human contributors pull back as AI spreads on the platform, there could be pressure to fill the gap with even more AI. In that case, people would be left mainly doing formatting and cleanup.

Another Wikipedian wrote that "the potential 'whittling down' each year of our pool of talented writers is the greatest threat to Wikipedia." That editor said the project needs more editors, especially experts, to fix weak and inaccurate articles.

The Wikimedia Foundation has said that editors' commitment to "reliable encyclopedic knowledge is something AI cannot replace," according to the excerpt. The author argues that this is not the same as saying it will not be replaced. The excerpt also says AI agents give plausible answers that are often wrong. It asks how those errors could be caught if Wikipedia itself becomes corrupted by AI.

What to watch next

• Whether Wikipedia's pool of volunteer writers keeps shrinking, which one editor called the project's greatest threat.
• Whether readership keeps falling after the loss of more than 1.1 billion monthly visits since 2022 that Christophe Henner reported.
• How the March 2026 rule limiting AI to translation and minor proofreading holds up as AI tools spread across the platform.