Staff and students in Pakistan describe judgement under uncertain AI rules

Updated Sep 07, 2026

student voicefeedback

Rules do not apply themselves. Ayesha Afzal, Shahid Rafiq and Martin Oliver explore how teachers and students make practical ethical decisions about generative AI when guidance is uncertain.

What the study examined

The researchers purposively recruited 12 faculty members and 12 students from Education and English departments in Pakistan. Online interviews lasted 30–45 minutes and were analysed using reflexive thematic analysis.

The authors describe moral improvisation: situated judgement involving norms, relationships and the demands of a particular academic task. Participants discussed uneven expectations, fairness and emotional responsibility.

They wanted clearer guidance, concrete examples and opportunities for shared discussion. The authors also stress that more detailed rules cannot eliminate interpretation and negotiation. Policy remains part of the setting in which judgement occurs.

This review covers the publisher's abstract and accessible methods, findings, discussion and conclusion passages, not the complete paper. The qualitative sample does not establish UK prevalence or demonstrate that a specific policy change improves outcomes.

A practical discussion prompt

Our suggestion is to use an example assignment when discussing AI guidance with students and staff. Ask what each person understands to be permitted, what needs disclosure and where judgement is still required. Record disagreements without assuming that every difference reflects misconduct or a defective policy.

Use those discussions to decide whether wording, examples or teaching support need review. A comment summary can organise reported concerns, but cannot establish whether a rule is fair in every situation or determine the intent behind a student's actions.

The student comment analysis governance checklist can help plan how feedback reaches a responsible team. This is an editorial application, not a tested intervention from the paper.

FAQ

Do the authors expect clearer rules to remove uncertainty?

No. They recommend shared guidance while recognising that judgement remains necessary.

Is moral improvisation another term for rule-breaking?

No. The concept describes efforts to act responsibly within uncertain, sometimes competing expectations.

Reference

Afzal, A., Rafiq, S., & Oliver, M. (2026). Moral improvisation with generative AI in higher education: faculty and student experiences in Pakistan. Teaching in Higher Education. Paper source.

Correction, 7 September 2026: Added the sample, restored the authors' caution that policy cannot eliminate judgement and qualified practical and product claims.