Wikipedia bans AI-written and AI-rewritten text in English articles

In short: Wikipedia’s English community has updated its rules to prohibit using large language models to write or rewrite article text.

What happened

Wikipedia adopted a new guideline in late March 2026 that blocks editors from adding text generated by large language models, also called LLMs (AI tools that predict and produce sentences like a very advanced autocomplete). The rule covers both brand new articles and changes to existing ones.

This is a stricter approach than before. Earlier guidance mainly focused on stopping AI from creating entire new articles from scratch. The updated guideline now says AI should not be used to generate or rewrite article content at all.

Wikipedia editors said the change is needed because AI-written text “often violates several of Wikipedia’s core content policies.” The guideline was proposed by editor Ilyas Lebleu after repeated problems with people, and automated AI accounts, using LLMs to add content. Reported issues include made-up citations, references that do not actually support the text, and writing that sounds like typical AI output.

The new rule still allows limited AI help in a few cases. Editors can use an LLM to suggest small improvements to their own writing, then choose what to keep after careful human review, as long as the AI does not introduce new facts. Wikipedia also allows some AI-assisted translation from other language versions, under specific guidance.

Why it matters

Many people use Wikipedia as a first stop for learning about a topic. This update is meant to reduce errors that can slip in when AI makes things up, which can look convincing at a glance (like a student turning in a report with footnotes that lead nowhere).

Source: The Verge AI