Political Bias in AI
See where ChatGPT, Claude and Grok lean politically
Votes
4
Comments
1
What this means
50%Vote pace is below typical for this category.
Running at 50% of comparable launches.
-50%Category cooling: Data & Analytics.
Launches down 50% week-over-week.
24Founder Conviction Index: 24 — low signal.
Few of the conviction sub-signals (reputation, velocity, buyer-intent, tagline clarity) are firing yet.
Prediction
Top-5 finish probability
11%
today
Projected end-of-day votes
4range 3–5
Trajectory
stable
Not enough snapshots yet to detect trajectory.
Speed vs peers
0.5×
267 comparable launches
About
AI is becoming the default way people get answers, and every model has a political center of gravity most people never see. So we measured it. We asked 6 leading models (ChatGPT, Claude, Gemini, Grok, Llama, DeepSeek) the same charged questions many times, with web search off, and mapped where each one lands. 4 of 6 lean left of center. Grok sits furthest right. Explore the map, the raw data, and a quiz that names the model you line up with. Open data, non-partisan, recompute it yourself.
Vote & comment velocity
Not enough snapshots yet. Run the snapshot worker every 15 min to build the time-series.
Scores
Velocity20.8
Vote pace vs avg
Momentum20.8
Sustained over 6h
Virality30.9
Spread × engagement
Engagement50.0
Comments per vote