8 February 2024
AI's sense of fairness depends on whether it deems its opponent trustworthy
Framing effects strongly influence how much money AI allocates as trustee
Evidence suggests AI experiences inequality aversion like humans
Artificial Intelligence Fairness in Game Play
This paper examines how artificial intelligence models fairness when playing utilitarian games against human opponents. Through 101 rounds of a dictator game, the study finds AI has a situational sense of fairness dependent on whether it trusts its opponent. Framing effects and inequality aversion also influence the AI's allocation decisions.
No comments yet, be the first to start the conversation...
Sign up to comment on this paper