Return to Article Details Political Bias in LLMs: Unaligned Moral Values in Agent-centric Simulations Download Download PDF