Answers are grounded in this paper and cite their sources.
Generative Artificial Intelligence (GenAI) can now produce assessment outputs across a wide range of disciplines on demand, calling into question the assumed link between a finished product and the learning that produced it. Institutional responses have largely focused on two approaches: detection, which is unreliable and inequitable, and permission or security frameworks, which specify how GenAI may be used but rarely how the quality of that use should be judged. Here, we argue that the defensible object of assessment is shifting from the product to the process, and more specifically to the student’s evaluative judgement when working with GenAI. Drawing on theories of evaluative judgement, self-regulated learning and authentic assessment, we propose the PACE framework, which comprises an iterative process pathway across five domains and a competency scale defined by levels of human agency, from deferring to the GenAI to directing and critiquing it. PACE decouples process from product in marking, requires evidence of evaluative judgement rather than relying on surveillance, and is proposed as a discipline-agnostic, equity-conscious complement to existing permission and security models. PACE offers a means to strengthen the validity of process-based evidence in conditions where the product may no longer be a proxy for learning.
No takes yet. Share an insight, caveat, or question.
Francis et al. (2026) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: