Rigor, commitment, and generative AI: a conceptual framework for student engagement in EAP writing

The increasing integration of GenAI into academic writing has prompted renewed attention to how students engage with writing tasks in English for Academic Purposes (EAP) contexts. While existing research has focused on GenAI's capabilities and risks, it has paid less attention to the instructional conditions that shape how students use these tools. This paper proposes the Rigor–Commitment Alignment Framework (RCAF), a conceptual model that examines the interaction between academic rigor and teacher commitment as two dimensions of the pedagogical environment. Drawing on research in academic literacies, engagement theory, self-determination theory, and visible learning, the framework identifies four instructional configurations (Neglectful, Indulgent, Punitive, and Rigor with Support) and links them, through a set of testable propositions, to four engagement profiles (deep-agentic, strategic, surface, and disengaged). The framework is grounded in Bandura’s (1986) triadic reciprocal model, in which personal factors, observable behavior, and the instructional environment reciprocally influence one another. The analysis suggests that patterns of GenAI use cannot be understood solely in terms of technological affordances, and may be substantially shaped by the interaction between task demands, teacher commitment, and student engagement. In this context, GenAI may either support higher-order academic literacy practices or facilitate low-effort task completion, depending on the alignment between rigor and commitment. The RCAF provides a conceptual basis for examining these dynamics and offers implications for EAP pedagogy and AI policy.