Wikipedia bans editors from using AI to generate article content

As artificial intelligence advances into the editorial and media landscapes, websites are working to establish rules for its use. This week, Wikipedia prohibited the use of AI-generated text by its editors, though it stopped short of completely banning the technology from its editorial processes.

In a recent policy update, the platform stated that “the use of LLMs to generate or rewrite article content is prohibited.”

This new wording clarifies and updates earlier, vaguer language, which stated that large language models (LLMs) “should not be used to generate new Wikipedia articles from scratch.”

The use of AI in Wikipedia articles has become a controversial issue among the site’s large, volunteer-based community of editors. According to 404 Media, the new policy, which was approved by a vote among the site’s editors, received an overwhelming majority support of 40 to 2.

However, the updated guidelines still permit limited AI use for certain editorial tasks.

“Editors are allowed to use large language models (LLMs) to suggest basic copy edits for their own writing,” the new policy states.

“They can incorporate some of these suggestions after conducting a thorough human review, as long as the LLM does not introduce any new content. Caution is necessary, as LLMs may go beyond the requests and alter the meaning of the text, potentially leading to mismatches with the cited sources.”