Risk-benefit perception of AI use : Public perception of AI in healthcare and commercial domains

Abstract: AI applications are today implemented in a wide range of settings with the goal of achieving greater efficiency. However, these implementations can not be guaranteed to be free of risks. This study investigated how people perceive these risks and benefits, and whether there were any notable differences to be found between the domains in which they appear, in this case e-commerce/marketing and healthcare. Additionally, the relationship between the perceptions and individual positive and negative attitudes towards AI were examined by utilizing an affect heuristic framework. The findings showed that the two domains did differ from one another, as ratings of both perceived risks and benefits were higher for the healthcare domain compared to the e-commerce/marketing domain. Further, an inverse correlation between ratings of risks and benefits were found within each domain, which is consistent with the affect heuristic framework.