The world’s largest online information resource, Wikipedia, has moved to prohibit articles generated by artificial intelligence. According to the updated guidelines finalized last weekend, editors are no longer permitted to use AI to write or rewrite articles. The primary driver for this decision is the tendency of AI-generated text to violate Wikipedia’s fundamental policies regarding verifiability and original research.
AI Content Prohibited in English Wikipedia
This new regulation currently applies only to the English version of Wikipedia. However, the platform still allows AI use in highly restricted scenarios. Editors can utilize large language models (LLMs) for basic copyediting and refinements, provided they do not introduce any new information or content. Additionally, AI-assisted translation from other languages into English is permitted, but only if the editor is sufficiently fluent in both languages to manually verify the accuracy of the output.
The new policy, proposed by the administrator Chaotic Enby and passed with overwhelming community support, also includes a warning about false positives. It notes that some human writers may naturally possess a style similar to language models; therefore, stylistic markers alone are insufficient to justify an editor’s suspension without further evidence of policy violations.
Combatting “AI Slop” and the Cleanup Initiative
Wikipedia editors have been struggling for months against a rising tide of “AI slop”—poorly written, substance-free content. This has led to the adoption of a speedy deletion policy (specifically under criterion G15) for pages generated without human review. Furthermore, a dedicated initiative called WikiProject AI Cleanup was established to systematically identify, scrutinize, and remove AI-generated misinformation and hallucinations from existing articles.
This decision underscores Wikipedia’s commitment to human-led verification over automated efficiency. What is your take on Wikipedia’s AI ban? Should these platforms be exclusively human-written, or can AI be a reliable partner within strict boundaries? Share your thoughts in the comments!
Would you like me to research the WikiProject AI Cleanup noticeboard for the latest blocked accounts or explain the technical differences between human prose and LLM-generated text as identified by Wikipedia’s internal guides?
