ai-bias

Open benchmark makes AI models' political bias measurable

The Neutrality Project launches an open, reproducible benchmark measuring political bias across six axes in leading AI language models.