Hospitals are increasingly integrating Artificial Intelligence (AI) into their operations with the goal of streamlining healthcare delivery processes and improving efficiency. However, this integration also introduces the risk of medical errors caused by AI, which can present significant financial and reputational challenges for hospitals. Despite these risks, little is known about how AI-driven errors influence patient reactions toward hospitals or how hospitals can effectively manage such responses. To address these gaps, the current research examines 1) how patients respond to medical errors caused by AI compared to those caused by humans or human-AI collaboration, and 2) how providing explanations for errors and varying levels of human involvement in AI decision-making shape reactions to AI-driven medical errors. Results from three studies (n = 1,281) show that AI errors elicit more negative reactions compared to errors caused by humans or human-AI collaboration (Studies 1 and 2). Specifically, individuals attributed higher responsibility to the hospital, were more likely to file complaints and pursue legal action and were less likely to recommend the hospital when the error was caused by AI. While providing explanations did not significantly mitigate these negative responses (Study 2), greater human involvement in AI decision-making helped buffer the adverse effects of errors that involve AI (Study 3). These findings provide insights for hospitals on balancing innovation with maintaining public trust.
