A Critical Review of Critical Thinking in HCI Research About AI

The importance of critical thinking when using generative AI in personal, professional, and educational contexts is widely acknowledged by scholars, educators, and politicians. There are contradictory hypotheses regarding the influence of AI on critical thinking, and research thus far has produced mixed results. In this paper, we present a critical scoping review of 80 research papers from Human-Computer Interaction (HCI) that study critical thinking with or about AI, focusing on how “critical thinking” is defined, measured, and how the papers claim critical thinking to be influenced by AI, according to current HCI research. Our findings demonstrate that critical thinking is rarely defined in HCI research about AI, and there are fundamental construct and criterion validity challenges in this line of research. We present a research agenda for more theory-grounded empirical research in critical thinking with or about AI, based on cognitive-psychological, philosophical, or educational research traditions.