Welcome and introduction Dr. Ouassim Karrakchou (from International University of Rabat and Local Organising Committee Chair)

Prof Therese Keane  (IPC chair)

Professor Don Passey (TC3 Chair) 

Prof Kaśka Porayska-Pomsta As generative AI gains momentum in education, public discourse increasingly positions education as the system in need of urgent reform, while presenting AI as a mature and inevitable solution. This keynote challenges that narrative by drawing on over five decades of research in Artificial Intelligence in Education (AIED), arguing that many of today’s headline concerns such as personalisation, scalability, and efficiency, are far from new, and that AIED’s accumulated insights are being overlooked just when they are most needed. Rather than casting AI as a replacement for teachers or assessments, I will highlight AI’s strength as a scientific instrument for investigating how people learn and the conditions under which learning thrives. This reframing invites deeper, often neglected questions, not only about what AI can do, but how its presence reshapes learners’ motivation, attention, memory, and agency, and what it therefore ought to do in order to support education meaningfully. By revisiting AIED’s core contributions – from instructional design and learner modelling to metacognition and socially grounded interaction – the talk offers a starting point for better understanding and anticipating what learners, educators, and systems should expect from AI. It explores the idea that the limitations of general-purpose AI systems are not only technical, but also conceptual and ethical: what we fail to ask about learning, we inevitably fail to design for. The keynote calls for a more reflective and historically informed conversation – one that positions AI first as a tool for deepening our understanding of learning and of the human capacities for education to cultivate, before it becomes a vehicle for desirable change.

Prof Arnold Pears: Since the release of ChatGPT, one of the most highly visible Large Language Model based ChatBots, in 2022 research has been prolific in all aspects of the educational research sphere. Generative AI models are in wide use in many industries, and prevalent use among student populations worldwide is a fact that Higher Education Institutions must accept. Research on generative AI initially dealt with the ability of the models to respond to assessment items in a range of disciplines. LLM systems are (due to their structure and the nature of the training data) inherently highly suited to formulating responses to many assessment questions, be they computer programming code, image generation, music, or more general purpose text. It became clear that these types of tools can generate very good responses to many assessment items, including examinations which are heavily based on recall of information. This is not surprising considering that the tools are trained on precisely the data that assessment items are asking to be recalled. There is also a growing literature on the use of generative AI in teaching and learning situations, ranging from using the tools to write assignment solutions to potential use as an interactive study partner. Use of variants of LLM trained on special purpose data-sets and integrated into other systems and tools have rapidly infiltrated all elements of academic work from teaching and learning to research. Generative AI tools are readily available to automate large components of data analysis (both qualitative and quantitative) as well as Grant Writing, academic article writing, and creating of social media and other types of communication. Indeed, academic staff are also using AI to automate aspects of the assessment process, provide automated tutoring to students, and to generate material for use in teaching.  In this talk I will argue that learning, and indeed teaching, should be intensely human activities in order for them to retain their social purpose and desired outcomes. Abdication of cognition to these powerful tools is a considerable risk facing our profession, and it is vital that we start to ask serious questions about when and how these tools are appropriate in the context of the academy.

Prof. Ernesto Cuadros-Vargas: Latin America remains one of computing education’s most overlooked regions—a land of untapped potential and unexpected innovations. This keynote demystifies CS education in Peru, Colombia, and Brazil, revealing how these countries tackle universal challenges with locally tailored solutions. In Peru, grassroots advocacy and accreditation reforms clash with infrastructure gaps, yet pioneering universities drive change. Colombia blends K-12 CS integration with bold university-industry partnerships, while Brazil leverages its research muscle to align curricula with global standards—yet all three face stark inequities between urban and rural access.vBeyond problems, the region offers lessons: agile policy experiments, hybrid pedagogies for resource-limited settings, and a culture of collaboration over competition. The talk closes with a call to action—for research partnerships, faculty exchanges, and a rethink of what "global" CS education truly means.

Don Passey (Lead Editor Post Conference Book) and Katherine Moretti (Springer Editor) - Special Issue of Education and Information Technologies.