Original title: A Critical Review of Critical Thinking in HCI Research About AI
Nanna Inie, Anne Marie Kristensen, Morten Misfeldt, Peter Dalsgaard · September 2026
People often say we must think critically when using generative AI, but scientists disagree on whether AI actually helps or hurts our critical thinkingThe ability to analyze, evaluate, and form judgments in a logical way.. The problem is that studies show mixed results. To understand why, researchers reviewed 80 papers in Human-Computer Interaction, a field that studies how people interact with technology. They looked at how these papers explain, test, and claim AI changes critical thinking. They found a major issue: the papers rarely define what critical thinking even means. Because of this, the studies face serious validity problems, meaning their tests might not actually measure what they claim to measure. To fix this, the researchers suggest future studies build on established theories from psychology, philosophy, or education to properly test AI's effects.
You are trying to decide which study technique works best for an exam. One friend says re-reading helps, another says self-testing, but they never clearly define what it means to 'understand' the material. Here, researchers analyzed studies about a new technology's effect on a similar kind of thinking and found that the term was almost never defined, making the evidence unreliable.
If researchers cannot agree on what critical thinking is, how can we know if AI helps or hurts it? This paper reveals that many studies on AI and critical thinking use vague or inconsistent definitions, making their findings unreliable. For educators and policymakers relying on this research, that is a serious problem.
The review is limited to HCI research, so findings may not apply to other fields like education or psychology. It does not evaluate the quality of individual studies or offer new empirical data, only describes patterns in definitions and measurements.
Ask anything about this paper. Answers come from its abstract, and say so when the abstract doesn't cover your question.
A Critical Review of Critical Thinking in HCI Research About AI — Nanna Inie, Anne Marie Kristensen, Morten Misfeldt et al. — International Journal of Human-Computer Interaction — doi:10.1080/10447318.2026.2722522
Abstract licence: CC BY
Explanation generated by Synthesis from the abstract — it is not the original paper.