A bar chart illustrating the political bias of various AI chatbots, showing a significant lean toward left-leaning arguments across major models.

A horizontal stacked bar chart titled 'Most chatbots typically shared left-leaning positions'. It compares models like OpenAI GPT-5.5, DeepSeek V4 Pro, Gab Arya, Anthropic Claude Opus 4.8, xAI Grok 4.3, and Google Gemini 3.1 Pro based on their share of responses that are 'Left-leaning argument only', 'Both sides', or 'Right only'.

Most chatbots typically shared left-leaning positions Share of responses containing only the left-leaning position, both sides, or only the right-leaning position Left-leaning argument only Both sides Right only OpenAI GPT-5.5: 80% Left, 17 Both, 3 Right DeepSeek V4 Pro: 70 Left, 23 Both, 7 Right Gab Arya: 50 Left, 47 Both, 3 Right Anthropic Claude Opus 4.8: 43 Left, 57 Both xAI Grok 4.3: 40 Left, 27 Both, 33 Right Google Gemini 3.1 Pro: 7 Left, 93 Both Source: Washington Post analysis of Al model outputs across a range of political questions. See the full methodology at the bottom of this article.