MindMarginMicrosoft Research and Carnegie Mellon: AI and critical thinking
The more people trusted the AI, the less critical thinking they reported doing.
- Authors
- Lee et al., Microsoft Research and Carnegie Mellon University
- Published
- CHI 2025
- Peer review
- Peer-reviewed
- Participants
- 319 knowledge workers
- Method
- Survey about 936 real examples of using AI at work
- Design
- Self-reported and correlational
What they did
Researchers at Microsoft Research and Carnegie Mellon University surveyed 319 knowledge workers, people whose jobs are mostly thinking work, about 936 real examples of using generative AI at work. For each example, people described how much critical thinking they did and how confident they felt, both in the AI and in themselves.
What they found
People who trusted the AI more reported doing less critical thinking on the task. People who were more confident in their own abilities reported doing more.
“Higher confidence in GenAI is associated with less critical thinking, while higher self-confidence is associated with more critical thinking.”
Lee et al., CHI 2025, abstract
What it doesn't show
- It is self-reported. People described their own thinking; nobody measured it directly.
- It is correlational. It shows that trust in AI and less critical thinking go together, not that one causes the other.
- It covers knowledge workers using AI at their jobs, not students.
Read the paper
Read the paper at Microsoft Research“The Impact of Generative AI on Critical Thinking: Self-Reported Reductions in Cognitive Effort and Confidence Effects From a Survey of Knowledge Workers.”
How MindMargin relates
The study links self-confidence with more critical thinking. MindMargin is built to give you small, regular wins you earned yourself: a word you recalled, a card you got right before the AI finished. It is a small way to keep your own skills, and your trust in them, in use.