ChatGPT in Economic Games: Experimenting with Human-AI Interaction

This study explores the interaction between artificial intelligence (AI) and human behavior in economic contexts, focusing on using generative AI, such as ChatGPT, to examine trust and reciprocity. Through an experiment conducted at LUMSA in Palermo, 61 economics students enrolled in a blended learning course participated in three economic games: the Ultimatum Game, the Trust Game, and the Public Good Game. ChatGPT tended to favor fairness and altruism, highlighting a difference in behavior compared to students who took a more individualistic approach. These results indicate the potential of ChatGPT as an educational tool to develop a critical understanding of ethical and economic dilemmas and suggest that interaction with AI systems can stimulate pro-social dynamics. Structuring prompts to ensure standardized and controlled results is also highlighted. The results of this work suggest future lines of research on human-computer interaction and the role of AI design in social and educational contexts.