AR
arXiv CS.AI
7/2/2026

Constructing Epistemic AI Literacy: Detecting Epistemic Aims and Processes in Student-AI Co-Programming
Short summary
Study of 10k+ student-AI interactions finds 78.8% use poor strategies like outsourcing, while only 11.1% show advanced epistemic engagement. Introduces Epistemic AI Literacy framework measuring how students think critically about AI-generated code.
- •78.8% of student-GenAI interactions rely on non-mastery-oriented outsourcing strategies
- •Only 11.1% demonstrate high epistemic engagement with verification and justification
- •New EAIL framework operationalizes observable dimensions of critical thinking in co-programming
Generated with AI, which can make mistakes.
Is this a good recommendation for you?



