The Impact of Generative AI on Critical Thinking: Self-Reported Reductions in Cognitive Effort and Confidence Effects From a Survey of Knowledge Workers
Citation. Lee, H.-P. (H.), Sarkar, A., Tankelevitch, L., Drosos, I., Rintel, S., Banks, R., & Wilson, N. (2025). The Impact of Generative AI on Critical Thinking: Self-Reported Reductions in Cognitive Effort and Confidence Effects From a Survey of Knowledge Workers. CHI Conference on Human Factors in Computing Systems (CHI ‘25). https://doi.org/10.1145/3706598.3713778
Summary. This CHI 2025 survey study examines how 319 knowledge workers perceive critical thinking when using GenAI tools at work, drawing on 936 first-hand examples. The authors find that higher task-specific confidence in GenAI is associated with less critical thinking, while higher self-confidence is associated with more critical thinking. GenAI appears to shift cognitive effort away from material production and toward information verification, response integration, and task stewardship.
Key claims
- Across 936 GenAI use examples, participants reported enacting some critical thinking activity in about 60% of cases (555/936).
- Confidence in AI doing the task negatively correlated with perceived critical thinking enaction (β=-0.69, p<0.001), while confidence in doing the task themselves (β=0.26, p=0.026) and evaluating AI responses (β=0.31, p=0.046) positively correlated with critical thinking enaction.
- Participants reported decreased effort for critical-thinking-related activities in most examples: 72% for Knowledge, 79% for Comprehension, 69% for Application, 72% for Analysis, 76% for Synthesis, and 55% for Evaluation.
- Higher confidence in AI was associated with less perceived effort for Knowledge, Comprehension, Analysis, Synthesis, and Evaluation; higher self-confidence was associated with more perceived effort in Application and Evaluation.
- Qualitatively, GenAI shifts critical thinking from information gathering to information verification, from problem-solving to response integration, and from task execution to task stewardship.
- Critical thinking was motivated by work quality, avoidance of negative outcomes, and skill development; inhibitors included low awareness, time pressure or job-scope constraints, reliance on AI, and difficulty improving outputs in unfamiliar domains.
- The authors argue that GenAI tools should support awareness, motivation, and ability for critical thinking rather than merely reducing cognitive effort.
Evidence & limitations
- Design: survey of 319 English-fluent knowledge workers who used GenAI tools at work at least once per week; analysis of 936 self-reported real-world GenAI use examples with quantitative models and qualitative coding.
- Strength: directly studies critical thinking in workplace GenAI use rather than relying only on output diversity or lab-task proxies.
- Limitations: self-report design; participants may conflate reduced effort using GenAI with reduced effort thinking critically with GenAI; subjective confidence may not be calibrated with objective expertise; English-only sample skewed younger and more technically skilled; cross-sectional design does not establish causality or long-term skill change.