Explicating multiple dimensions of public’s and technology developers’ mental models of risks and benefits of artificial intelligence in autonomous vehicles in Singapore

• Technological attributes and implications of use are key themes in mental models. • Environmental, economic, personal, and societal dimensions of implications of use. • Drivers and non-drivers draw on everyday experiences to develop mental models. • Domain-specific knowledge explains difference between public and expert mental models. • Availability of ride-hailing services is a common basis for development of mental models. Developments of artificial intelligence for autonomous vehicles promise better commuting experiences. However, current limitations of artificial intelligence signal some risks associated with autonomous vehicle use. Based on the trajectory of artificial intelligence development, technology developers and the public would develop mental models of risks and benefits of artificial intelligence in autonomous vehicles. Mapping drivers’, non-drivers’, and technology developers’ mental models, this study found two broad themes: technological attributes and implications of autonomous vehicle use . Specifically, there are environmental , economic , personal , and societal dimensions of implications. Benefits associated with prospective introductions of autonomous vehicle-enabled transport services is a key similarity among all mental models. Notable differences occur between the technological and societal dimensions of mental models of risks. The results are based on seven focus group discussions with drivers, non-drivers, and technology developers in Singapore.