Wikipedia Bans AI for Generating Article Content

Wikipedia has implemented a new policy prohibiting editors from using artificial intelligence tools, specifically large language models, to generate or rewrite article content across its platforms. The update, announced recently, aims to uphold the site's standards for accuracy and reliability, allowing AI only in limited cases such as basic copyediting and translation, both requiring human oversight.

The policy explicitly states that text produced by large language models often violates Wikipedia's core content guidelines, including requirements for verifiability, neutrality and the prohibition on original research. All material on the site must be attributable to reliable, published sources, and editors cannot introduce new analysis or synthesis not clearly supported by those sources. AI-generated content frequently hallucinates facts, fabricates sources or alters meanings in ways that undermine these principles. "For this reason, the use of LLMs to generate or rewrite article content is prohibited," the guidance reads, emphasizing that human judgment remains essential.

Exceptions are narrowly defined to prevent misuse. Editors may use AI to suggest basic copyedits on their...