Post-instrumental assessment: certify the learner's accountable relation to AI-mediated work via a design sequence of naming the capacity, classifying AI use, and requiring evidence at the point of judgment¶

Strategy · All strategies
Evidence · no claims cited

Description¶

This strategy implements post-instrumental assessment through a three-step design sequence: "First, name the target capacity", then decide which forms of AI use would scaffold versus replace that capacity, then require evidence at the point where judgment matters, such as a justification of source selection, a defense of a contested premise, or an explanation of why an AI suggestion was rejected. Documentation should be lean and meaningful, focused on responsibility rather than exhaustive prompt logs, because a short process note, targeted annotation, viva, or in-class transfer task asks the student to reason under a new demand. The strategy aligns the assessment rule with the capacity the assessment claims to certify rather than applying a universal AI ban or permission.

Design Implications¶

Context¶

Requirements¶

• The institution must be able to state what capacity the assessment is meant to evidence before choosing enforcement measures

Constraints¶

• Equity checks are needed: policies treating all AI use as suspicious may punish disabled students, multilingual writers, and students using legitimate support

Target Learners¶

• higher education and school students
• instructors and assessment designers

Target Learning Goals¶

• demonstrating an accountable relation to AI-mediated work, including framing tasks, evaluating evidence, revising claims, and answering objections

Affordances¶

• Post Instrumental Learning

Related Strategies¶

• Formative friction: require goal statements, alternative comparison, and justification before high-stakes delegation, kept proportionate and accessible
• Guide students on when conversational AI supports learning versus when it replaces checking, evaluation, and independent judgment

Examples¶

-

Key Sources¶

• Kai Yao. (2026). When AI Does the Work, What Is Learning For? Post-Instrumental Learning and the Risk of Capacity Dissolution. arXiv. https://arxiv.org/abs/2607.28041