Co-hosted by the Global MOOC and Online Education Alliance (GMA), Tsinghua University, XuetangX, and the UNESCO Institute for Information Technologies in Education (IITE), the second session of the 2026 Online Education Dialogue (OED), under the theme of “Educator, Curriculum, Pedagogy,” was successfully held.
The session featured distinguished speakers: Professor Darkhan Bilyalov (Former Vice President of Nazarbayev University), Dr Jenae Cohn (Director, Center for Teaching and Learning, UC Berkeley), and Professor Michael Phillips (Professor of Digital Transformation, Faculty of Education, Monash University), who shared expert perspectives on the theme of “Educator, Curriculum, Pedagogy”. The dialogue was hosted by Li Yifan, Senior Manager, Online Education Center, Tsinghua University, and Assistant Secretary-General, Global MOOC and Online Education Alliance.
(From left to right) First row: Prof. Darkhan Bilyalov, Prof. Michael Phillips
Second row: Dr. Jenae Cohn, Li Yifan
Professor Darkhan Bilyalov argued that higher education needs a more systematic approach to AI integration across university programs, as many institutions are still engaging with AI in fragmented and uneven ways. Drawing on a five-level model for structuring AI adoption, he outlined a pathway ranging from treating AI as an object of study to embedding it into capstone projects and broader academic practice. He suggested that the challenge is not only how to incorporate AI into existing systems, but how to move beyond sporadic experimentation toward more coherent institutional frameworks for teaching and learning. Looking ahead, he emphasized that as AI continues to expand its capabilities, universities will need to reconsider conventional teaching structures and assessment models, and that educational authority in the future will depend less on speed or informational advantage and more on the ability to frame questions, design learning experiences, and cultivate sound judgment.
Dr Jenae Cohn situated generative AI within the practical realities of teaching and learning, emphasizing that effective integration begins with clear, transparent, and discussable policies embedded in course design. Drawing on UC Berkeley’s work, she argued that AI guidance should not remain as abstract statements in syllabi, but should be translated into visible and intelligible support for students across assignments and learning activities. She also highlighted the importance of openness and vulnerability on the part of educators, suggesting that conversations about AI should help students better understand learning goals, academic integrity, and ethical participation, rather than rely solely on prohibition or compliance. More broadly, she underscored the value of co-creating norms with students and ensuring that assessment continues to support meaningful learning rather than surface-level performance.
Professor Michael Phillips examined the pedagogical implications of AI use in educational settings, particularly through the lens of faculty and student experience. Drawing on discussions at Monash University, he observed that many students experience ambiguity, discomfort, or even guilt when using AI, especially where expectations and boundaries are not clearly communicated. He reflected on the idea of “cognitive offloading,” suggesting that while AI may appropriately assist with routine or low-stakes tasks, deeper reasoning and complex intellectual work should remain central to student learning. He concluded by emphasizing the enduring importance of professional judgment in teaching, arguing that teacher authority remains indispensable not because educators can compete with AI on speed, but because they are able to interpret context, respond to student needs, and exercise nuanced pedagogical judgment in ways that technology alone cannot replicate.
Panel Discussion
In the panel discussion moderated by Li Yifan, the speakers explored several key questions concerning educators, curriculum, and pedagogy in an AI-rich era. Bilyalov noted that as students increasingly engage with systems that appear faster, broader, and more responsive than individual instructors, the meaning of teaching authority may need to be reconsidered. Cohn emphasized the importance of openness, ethical clarity, and transparent communication with students about learning goals and responsible AI use. Phillips added that teacher authority remains important because educators are able to interpret context, respond to student needs, and exercise pedagogical judgment in complex learning environments.
The discussion also turned to curriculum reform. The panelists considered not only what AI-related knowledge and competencies should be added, but also what may need to be reduced, revised, or removed in an AI age. They noted that curriculum redesign should be guided by learning objectives rather than technological novelty, and discussed whether curriculum should continue to be understood primarily as a fixed progression or move toward a more navigational model of learning.
Another major thread concerned the boundary between AI as a thought partner and AI as a substitute for thinking. The speakers agreed that this distinction should be addressed carefully through pedagogy and assessment. Phillips described approaches in which AI can provide suggestions without displacing student reasoning, while Bilyalov stressed the importance of guardrails and assessment mechanisms that ensure continued student engagement. Cohn emphasized the value of reasonable regulation and a learning culture that continues to prize independence, reflection, and ethical judgment.
Finally, the panel highlighted the importance of integrating AI into higher education in ways that remain human-centered, evidence-based, and pedagogically grounded. The dialogue underscored that the challenge is not only how to introduce AI into existing systems, but also how to rethink the role of educators, curriculum, and pedagogy in response to a rapidly evolving learning environment.