The rapid expansion of generative artificial intelligence (GenAI) has blurred the boundaries between automation and authentic learning, intensifying the tension between AI‑assisted output and evidence of human cognition. While GenAI can automate writing and problem‑solving, academic integrity requires clarity about what constitutes evidence of learning, judgement, and capability. Despite regulatory guidance from TEQSA (2024) to mitigate emerging risks, educators continue to face ambiguity in communicating what counts as legitimate evidence of learning in AI‑enabled environments.

Current practices focus largely on AI detection, plagiarism risk, and policy compliance, offering limited guidance to demonstrate authentic cognitive engagement with systemic communication of process‑level transparency (reflection, attribution, or interaction logging) that distinguishes human reasoning from automated output. Assessment design literature also lacks frameworks that integrate AI use while still requiring to capture metacognition, disciplinary thinking, and evaluative judgement.

This poster addresses these communication challenges by proposing a coherent, design‑driven model for academic integrity in AI‑enabled education. It offers practical strategies for making human reasoning more visible, traceable, and assessable through structured reflection, transparent attribution, and redesigned assessment workflows. By reframing integrity as a design and dialogue problem rather than a compliance exercise, the model shifts institutions from punitive detection toward proactive communication practices and shared responsibility.

TEQSA (2024) Gen AI strategies for Australian higher education: Emerging practice shape integrity culture by making support recognisable, memorable and present across the student experience.