Use a six-question drafting and audit framework for GenAI assessment rules covering task rules, defaults, disclosure, contribution, verification, and hierarchy¶

Strategy · All strategies
Evidence · 1 claim (1 for) · 1 study (1 qualitative), q2 · 0 of 1 report an effect size · 1 claim rests on one study

Description¶

The article proposes a six-question framework institutions can use to draft and audit GenAI assessment rules: task rule, default and authority, disclosure, contribution and authorship, verification, and hierarchy and escalation. The article states that "Institutions may choose different substantive defaults, but their written environments should answer each question", and that any unanswered question marks a predictable source of uncertainty. The framework starts with task instructions because they produced the clearest result in the vignette study. Regulators can add policy-vignette stress tests to provider toolkits.

Design Implications¶

Context¶

Requirements¶

• Each written policy environment must answer all six questions, including stating a fallback rule and naming the approving role

Constraints¶

• The framework does not prescribe a single governance architecture; it tests whether the chosen model remains coherent when task instructions are silent, in conflict, or expressly prohibit use

Target Learners¶

• University students
• Educators
• Academic integrity decision-makers

Target Learning Goals¶

• Recovering a defensible classification of GenAI use from written policy

Affordances¶

• Operational Classification Model Genai Governance

Claims¶

• Students express profound confusion about whether GenAI-assisted coursework is original work and whether they can claim authorship [+W]

Related Strategies¶

• Replace detector-centred assessment with a four-stage process-oriented model: declared starting point, versioned drafting, process artefacts, and targeted oral verification
• Four-layer workplace assessment battery with non-compensatory decision rules for competent GenAI use
• Provide explicit instructional boundary cues, process-oriented assessment, and peer modeling of verification to frame responsible GenAI use

Examples¶

-

Key Sources¶

• Biranchi Poudyal. (2026). Mapping the Authorized Boundary: A Comparative Policy-Vignette Study of Generative AI Governance in Australian Higher Education. https://orcid.org/0000-0002-7210-5480