Unlikely allies: ChatGPT and higher education assessment

Abstract

The emergence and spread of generative AI, including Large Language Models (LLMs) like ChatGPT is raising uncomfortable questions about the nature of assessment in higher education. Learners placed on the right side of the ‘technology divide’ can produce rapid essays and reports, formatted to any given specification and written with native fluency. Non-traditional learners can overcome the hurdle of academic writing, without recourse to pernicious ‘essay mills’. ChatGPT can ease the anxiety of assessment ‘bottlenecks’, during periods where students need to complete multiple submissions over a short period of time. So, does the increasing presence of generative AI in higher education, indicate assessments becoming a ‘vestigial’ formality? Do AI shortcuts signal the need to roll back time and, returning to on campus exams, once again test students’ recollection, rather than learning? In this opinion piece we reflect on our experience as higher education lecturers and suggest ways in which generative AI can become an assessment ally in the name of impactful learning.