Educational AI, learning and skill
Learning is the slow work of building a mind that can manage on its own. AI can help with that work, or it can do the work for us. A tutor that asks a good question at the right moment can lift a struggling student. A tool that hands over finished answers can leave the same student feeling capable while learning little. The difference often lies in design, and in how teachers, parents and learners choose to use these tools. The stakes are practical, because exams, jobs and health care depend on real skill. They are also personal, because the confidence that comes from earning a skill is hard to replace.
Wonder
A patient teacher for all
Every learner has a patient tutor, and every teacher has help.
Over double
the median test gains with a well-designed AI tutor, against an active-learning class, in a randomized trial of 194 Harvard physics students. Measured
4 points
more likely to master a topic: students whose human tutors had an AI assistant. The gain was 9 points with lower-rated tutors. Not yet peer reviewed. Measured
0.31 standard deviations
higher scores for Nigerian secondary students after six weeks of teacher-guided sessions with an AI tutor, in a World Bank randomized trial. Only 759 of 1,328 students sat the final test, and the study is an unreviewed working paper. Measured
66.2% vs 60.7%
Students tutored by an AI that human tutors supervised solved a new problem on the next topic slightly more often than with human tutors alone. The difference's uncertainty range includes zero, and the study is a preprint by the AI's developers. Measured
It takes Guardrails. In another trial a plain chatbot left students worse off once it was taken away.
Worry
Learning and skill
+48%, then 17% worse
Students using plain GPT-4 for math practice scored 48% better in practice, then 17% worse than classmates on an exam without AI. A tutor with guardrails largely avoided the harm. Measured
28.4% to 22.4%
Experienced doctors' adenoma detection rate in colonoscopies done without AI, before and after they began using AI assistance. Observational. Measured
67%
of US students agree that the more students use AI for schoolwork, the more it will harm critical thinking. Survey
12%
of UK undergraduates said in a December 2025 survey that they had put AI-generated text directly into assessed work, up from 8% in 2025 and 3% in 2024. These are self-reports. Survey
59%
of US teens said using AI to cheat happens at least somewhat often at their school, and 10% said they do all or most of their schoolwork with a chatbot's help. The first figure is what teens perceive. Survey
Less critical thinking
was reported by knowledge workers who had more confidence in AI, in a survey of 319 workers describing 936 tasks. This is a self-reported association and does not show cause. Survey
What is being done
Schools are being required to decide.
Ohio law required every school district, community school and STEM school to adopt a policy on AI use by 1 July 2026. Schools could adopt the state's model policy or write their own.
For artists and musicians
Nobody has measured it for the arts.
No controlled study has yet measured how much skill is lost, or gained, when musicians, painters or writers lean on AI. The evidence so far comes from math classrooms and hospital clinics.
Worth knowing
A much-cited positive result was withdrawn.
A widely cited 2025 meta-analysis reporting large learning gains from ChatGPT was retracted in April 2026. The editor cited discrepancies in the analysis. The case for learning gains rests on fewer and shorter trials than the headlines suggest.
An open question
Do skills learned with AI support last for years once the support is taken away, or do they fade?
This page does not list everything. Last reviewed October 11, 2026.
- Measured counted or experimental data
- Survey what people report or believe
- Estimate a judgment or model, often a company's own
- Forecast a projection of the future
- Reported a company's, agency's or journalist's own account