What should we be assessing exactly? Higher education staff narratives on gen AI integration of assessment in a postplagiarism era

Generative artificial intelligence (GenAI) has challenged assumptions about assessments in higher education. Although GenAI chatbots create opportunities for inclusive learning, their misuse raises new concerns about academic integrity. This study provides empirical insight to inform assessment practices designed to include chatbots ethically. Guided by a postplagiarism framing, our research question was: How do educators make meaning of the ethical boundaries involved in evaluating students’ skills when chatbots are integrated in assessments? Adopting a qualitative design informed by an interpretive paradigm and narrative inquiry, we conducted a thematic narrative analysis of interviews with higher education staff (N = 28). Educators positioned themselves as stewards of learning with integrity within their disciplines, delineating boundaries that distinguished acceptable from unacceptable chatbot use. Across narratives, participants viewed prompting and critical thinking skills as compatible with chatbot integration. In contrast, perspectives on assessments focusing on writing skills were nuanced, ranging from specific inclusion to complete exclusion, depending on intended learning outcomes, disciplinary conventions and individual conceptions about learning with integrity. Educators also identified broader challenges, such as the potential for language standardisation, overreliance, blurred authorship and untraceable forms of cheating. These findings extend ongoing debates about integrity and assessments in technologically mediated learning.