AI Acceptance Model for Public Value Creation: Regulatory Perception with the Moderating Effects of Social Expectancy Disconfirmation

As artificial intelligence (AI) expands into a core infrastructure of public administration, the imperatives of accelerating innovation and imposing precautionary regulation coexist, yet their combined influence on citizens’ willingness to adopt AI remains largely unexamined. This study brings together three elements—regulatory perception, social expectation disconfirmation, and adoption intention—within a single analytical framework to explain how concurrent calls for AI innovation and strict oversight shape public acceptance. In doing so, it connects research traditions that have often been separate: technology acceptance models focused on functional benefits, risk studies centered on regulation, and expectation disconfirmation work that treats satisfaction change as one directional. The findings show that stronger support for regulating AI lowers willingness to adopt, whereas negative social expectation disconfirmation—the feeling that AI falls short of social expectations—paradoxically heightens citizens’ desire for further adoption and improvement. When these two factors combine, regulatory support magnifies adoption resistance, revealing a multi layered interaction among regulation, perceived risk, and acceptance. The results demonstrate that expectation disconfirmation can generate not only technological resistance but also pressure for innovation. They also indicate that effective AI governance should pair transparent risk management with participatory improvement platforms to maximize public value.