Moral Mimicry: Large Language Models Produce Moral Rationalizations Tailored to Political Identity

This paper studies whether LLMs moral preferences based on prompted political ideology replicate known results obtained in social science studies, using tools from Moral Foundations Theory