Wikipedia Says No to AI-Generated Text in Articles but Makes Two Exceptions

In recent years, Wikipedia reformed its content policy by banning artificial intelligence (AI)-generated text in articles. This is a new rule that explicitly prohibits the use of large language models (LLMs) when writing an article or rewriting e-page for the website. The platform has also made two exceptions to editors, allowing them to use such tools for copyedits on pages and translate pages from any language to English, while taking strong anti-AI position. Nevertheless, it has warned contributors to be cautious when using AI chatbots.

Wikipedia Takes Strong Stance Against AI on Its Platform

During a new project page, Wikipedia posted the updated content policy ‘It is prohibited for LLMs to create or rewrite article content. In a statement on the open-source online encyclopedia, it was made as “using AI-generated text from chatbots like ChatGPT, Gemini, Claude, DeepSeek and others” that “violates many of Wikipedia’s core content policies. Paraphrast.

The non-profit platform’s main problem the text is trying to solve is its verifiability and neutrality; AI-generated content sometimes changes the meaning of the texts, so that it can not be supported by cited sources. A problem with the hallucination (along with Wikipedia’s reliance on quality of articles) is that it can also cause errors in accuracy, as this may be related to the AI-related halludance problem.

It has also made two exceptions to editors, though, as it has argued for paraphrasingr. editor) can use LLMs and chatbots first, where editors may suggest basic copyedits to their own writing. If the AI does not add content of its own, these can also be added to the page after human review. Wikimedia does ask editors to be careful when using such tools.

In addition, Wikipedia is also allowing editors to translate articles from another language to the English Wikipedia using AI chatbots for second-hand translation of their work. Yet editors have also been urged to follow the rules of translation for LLM-assisted. Essentially, editors must tag such text as “automatically translated” text needing review of it. These are approved only after human review of s.

This is a time when Wikipedia’s move comes amid growing social media sites with posts that are generated by generic AI, and many have expressed concerns about replacing the text of human-written texts as well as its authenticity.