Teach AI-mediated judgement and make process, judgement and transfer visible in assessment¶

Strategy · All strategies
Evidence · no claims cited

Description¶

The review recommends pedagogy that moves from policing AI use to teaching AI-mediated judgement: learners should practise lateral reading, source triangulation and annotate AI responses to identify unsupported assumptions and recover sources. Assessment should reward process and transfer through staged drafts, oral defence, source maps and occasions to perform core capabilities with reduced AI support, so that learning outcomes remain recoverable and transferable rather than dependence becoming predictable.

Design Implications¶

Context¶

Requirements¶

• Feedback tasks should require comparison between AI, peer, teacher and exemplar-based feedback, followed by a justified revision decision

Constraints¶

• Such assessment designs should not become surveillance-heavy or disadvantage students who use assistive technologies

Target Learners¶

• Students and teachers in higher education and schools using generative AI

Target Learning Goals¶

• Evaluative judgement, feedback literacy, verification practice and transfer of capability under reduced AI support

Affordances¶

• Relational Epistemic Agency Norm

Related Strategies¶

• Direct Instruction Web Based Reading Strategies
• Consider the News Source
• Design AI learning systems for inspectable, contestable use: provenance, represented uncertainty, user control and epistemic plurality
• Embed SRL-oriented AIGC pedagogy: prompt-and-reflect routines, verification rubrics, and feedback literacy
• Assessment design levers that make learning visible and reduce the payoff from generic generated text

Examples¶

-

Key Sources¶

• Du Y, Yuan Y (2026) Epistemic dependence in AI-mediated learning. AI & SOCIETY. https://doi.org/10.1007/s00146-026-03294-1