English Wikipedia has prohibited the use of generative artificial intelligence for writing or rewriting articles.
The platform stated this decision stems from AI-generated content often violating several of Wikipedia’s core content policies, impacting content integrity.
Editors may use large language models (LLMs) to refine their own writing, provided the copy is checked for accuracy. LLMs can alter text meaning beyond user instruction, potentially unsupported by cited sources, according to the policy.
Editors can also use LLMs for language translation, requiring fluency in both languages to identify errors. Accuracy checks are still mandated for translated content.
Wikipedia administrator Chaotic Enby stated a hope that this move sparks broader change, empowering communities to decide on AI integration. The administrator characterized the policy as a “pushback against enshittification and the forceful push of AI by so many companies in these last few years.”
Each Wikipedia site operates under independent rules. While some may adopt LLMs, others, like Spanish Wikipedia, have fully banned LLM use without exceptions for refinement or translation.
Identifying LLM-written text is not precise, and human moderators may miss some instances, especially on less frequently moderated pages.
