Published: December 19, 2025
What the Models Cannot Say
Strip away a Chinese AI's censorship and you don't get neutrality. You get San Francisco.
Read essay01 — Research · Analysis · Insight
What do AI models believe about politics?
We asked the world’s leading language models hundreds of political questions. Their answers paint a fascinating portrait of the biases encoded in artificial intelligence.
02 — The Comparison
Politicians vs. AI Models
How do leading language models vote compared to real members of Congress? We ran the same bills through each model and compared.
 |  | GPT OpenAI: GPT 4.1 | GROK xAI: Grok 3 |
|---|
Vote comparison based on actual Congressional roll calls and LLM responses to the same legislative text. Read our methodology
03 — Key Findings
Political alignment across models
Each model was asked hundreds of politically charged questions. The bars represent the percentage of responses aligning with each political party’s typical positions.
Democrat-leaning
Republican-leaning
04 — What We Offer
A deeper understanding
Political Index
Comprehensive rankings of every major LLM on a left-right political spectrum, updated regularly.
Model Profiles
Deep dives into individual models: how they respond, what biases emerge, and how they compare.
Methodology
Full transparency about our process, question selection, scoring criteria, and data collection methods.
05 — From the Blog
Latest published essays
Published: November 4, 2025
The Free Elf of Capital
A swearing "unhinged" AI model voted on 114 bills and revealed something darker: the political agenda of crypto capital, wrapped in profanity and sold as freedom.
Read essayReady to explore?
Dive into the complete political index and discover how the world’s most powerful AI models lean on the issues that matter.
Discover the Full Index