Generative AI and a Single Integrity Code for London and Online Haverton Students

Abstract

Two pathways, one award, one integrity code. This study compares detector flags with viva and process-note evidence in both cohorts. Detectors misfired on international English. Process evidence was fairer. Campus students were not ‘safer’ because they sat in a room if the task was take-home. Haverton University should ban pathway-split panic. Research here is policy: declare tools, require process, examine orally when the stake is the award.