As generative AI becomes embedded in education, the central question is shifting: not “Did AI write this?” but “Did meaningful learning take place?”. In this conversational talk, Dr Eduardo Oliveira (University of Melbourne) explores this question through the lens of Human-AI Hybrid Intelligence. Drawing on recent work in authorship verification and AI-powered code review, he highlights the risks of externalisation, where AI replaces critical learning processes. He then presents Melbourne’s initiatives in AI literacy, showing how AI can shape mental models when embedded in pedagogy. Finally, he looks forward to assured learning, a framework for extending human cognition through hybrid human-AI systems that make the process of learning visible and valued. The talk offers a vision for moving beyond fear of AI “cheating” towards designing futures where human-AI partnerships enhance trust, capability, and deep learning.