How generative artificial intelligence shapes citizens’ perceptions of public value: An experimental study

Generative artificial intelligence (GAI) represents a significant advance in AI technology, offering new possibilities for enhancing citizen-government interactions. This study explores the impact of GAI on citizens’ perceptions of public value creation within the context of government-citizen interactions, which has yet to be examined empirically. Utilizing two survey experiments with a representative sample of 2293 citizens, we compare GAI responses to two other government responses (selective-avoidant and fact-value) and further explore the moderating role of disclosing GAI usage on these relationships. Compared to selective-avoidant responses, GAI responses significantly enhance citizens’ perceptions of public value creation in terms of usefulness, fairness, trustworthiness, and emotional reaction. However, there was no significant difference between GAI and fact-value responses; both were rated highly in terms of public value creation. Importantly, disclosing GAI usage markedly diminishes perceived public value across all response types. In particular, the positive impact of GAI responses relative to selective-avoidant responses is reduced. Our findings illuminate the potential of integrating GAI into public services and underscore the complex dynamics around transparency.