Overview

• Wikipedia articles use more negative language for right-leaning public figures compared to left-leaning ones.
• Left-leaning terms are more often linked with positive emotions like joy, while right-leaning terms are associated with anger and disgust.
• The study suggests Wikipedia's bias could influence AI models trained on its content, including OpenAI's ChatGPT.
• Wikipedia's neutrality policy appears to be failing, according to the study's findings.
• The Manhattan Institute report calls for improved content standards to ensure unbiased information.