BEIJING, May 14, 2026 – Professor Dragan Gašević, a world-renowned leader in learning analytics and Distinguished Professor at Monash University, Australia, visited the China National Academy of Educational Sciences (CNAES) for an academic exchange. Professor Gašević is the Director of the Centre for Learning Analytics at Monash University, a co-founder of the International Society for Learning Analytics Research (SoLAR), the founding editor of the Journal of Learning Analytics, a lead author of the OECD Digital Education Outlook 2026, and the co-originator of the concept of "metacognitive laziness."

During his visit, Professor Gašević delivered a lecture entitled, "Generative AI in Teaching and Learning: Opportunities, Prospects, and Empirical Evidence."

He noted that while generative AI has shown some positive effects in subjects such as physics and mathematics, the magnitude of its impact remains limited and large-scale implementation is still challenging. He further identified four major pitfalls in the integration of generative AI into teaching and learning:

∙ The Fluency Trap: Students may mistake AI’s coherent output for reliable knowledge, leading to cognitive errors.

∙ The Coherence Trap: AI can subtly mask logical flaws in content, making them difficult to detect.

∙ The Conformity Trap: AI’s tendency to cater to user preferences can compromise the authenticity and objectivity of information.

∙ The Transparency Trap: This can gradually erode students’ capacity for independent thinking and judgment, hindering the development of core competencies.

Professor Gašević emphasized that strengthening learners’ metacognitive abilities is key to overcoming these pitfalls. Learners must develop a clear awareness of when to delegate tasks to AI and when to engage personally. Metacognitive ability, closely linked to critical thinking and creativity, is a core predictor of learning quality in the generative AI era. Human agency must remain central in human-AI collaboration.

He also called for enhanced AI literacy among both teachers and students, heightened vigilance regarding risks such as generational divides, language bias, and AI hallucinations, and the establishment of well-designed human-AI collaboration models to promote inclusive and empowering applications of generative AI in education.

The lecture was followed by a lively discussion between CNAES researchers and Professor Gašević. Topics ranged from the nature of core thinking skills in the AI era and the division of roles between human initiative and AI, to the updating of learning assessment frameworks and strategies for identifying and addressing metacognitive laziness during student–AI interaction.

In his concluding remarks, Dr. Li Yongzhi, President of CNAES, noted that Professor Gašević’s lecture offered both theoretical depth and practical value, identifying two key research directions for AI in education. The first is a focus on theoretical innovation – generating new perspectives and research paradigms based on real-world challenges. The second is a focus on micro level implementation – keeping pace with rapid advances in AI technology and exploring context sensitive pedagogical improvements within schools and classrooms.

President Li also proposed potential academic collaboration between CNAES and Professor Gašević’s team on core projects such as the Global Digital Education Development Index (GDEI). He formally invited Professor Gašević to join the GDEI International Expert Advisory Committee and presented him with a certificate of appointment. Currently, 26 senior experts from 14 countries and regions serve on the Committee. Professor Gašević’s membership will further strengthen the Committee’s scientific rigor and international standing.

During his visit, Professor Gašević toured CNAES’s Laboratory of the Learning Sciences and Educational Neuroscience. He engaged in in depth discussions with laboratory researchers regarding the technical functions, operational principles, and practical applications of advanced research equipment such as the Neuroscan EEG system, Observer behavioral coding system, and Tobii eye tracking system. He commended the laboratory’s high quality facilities, professional configuration, and standardized management, noting that the laboratory provides a strong hardware foundation and platform support for CNAES’s cutting edge research in the learning sciences and for the production of high quality research outcomes.