GPT, Gemini, and Grok Exhibit Clear Left-Leaning Political Bias
- Source
- Washington Post
- Time
- 11:32 AM
- Weight
- 94/100
A recent study conducted by The Washington Post indicates that several leading artificial intelligence chatbots exhibit a consistent left-leaning political bias. After testing models from OpenAI, Google, Anthropic, xAI, and others using a series of hot-button political questions, researchers found that the outputs frequently favored progressive viewpoints.
OpenAI’s GPT model was identified as having the most pronounced skew, answering nearly 80 percent of queries with exclusively left-leaning arguments on topics such as the electoral college, taxation, and healthcare. While Google’s Gemini generally adopted a "both-sides" approach by providing balanced perspectives, even models marketed as conservative or "anti-woke," such as Elon Musk’s Grok and Gab’s Arya, cited left-leaning arguments more frequently than right-leaning ones on average.