Require students to submit the prompts, AI outputs, and a short justification of what they accepted, rejected, or modified when AI helped produce coursework. The goal is to transform AI from an AnswerBot into a documented tutoring interaction that can be evaluated for reasoning and learning rather than just for outcomes.
— If adopted widely, this policy changes academic assessment, affects credentialing incentives, and sets precedent for transparency standards around human–AI collaboration.
Tyler Cowen
2026.10.30
85% relevant
Cowen’s proposal centers on grading students not on answers but on the workflow and AI logs they produce (a professor grades the student's AI log), which directly maps to the existing idea that students should be required to record and disclose AI use as part of assessment and academic policy.
Arnold Kling
2026.05.03
100% relevant
Mark McNeilly's recommendation that students document AI prompts, outputs, and their decision process when using AI for assignments.
← Back to all ideas