Teaching Together with Virtual Humans

Designing a Human-AI Collaboration Intervention with and for Online Teachers

Abstract

This short paper presents the design phase of an intervention study exploring human-AI collaboration with a virtual human in online teacher education. The study investigates how online K-12 teachers, through collaboration with a virtual human, develop AI literacy, ethical awareness, and pedagogical reflection within their online professional learning environment. The study, to be conducted at an online international school, adopts the ADDIE instructional design model and draws on the European AI Literacy Framework, a framework for teacher professional development, and principles of Networked Learning. The virtual human is a responsive, embodied AI agent powered by generative technology using existing virtual human technology. The virtual human acts as a co-instructor and reflective partner for participating online teachers, supporting them as they explore the educational use of AI and learn to create AI-generated visuals to enhance their online teaching practice. Research objectives include: (1) modelling human-AI collaboration through joint instruction by the researcher and the virtual human to support online teacher professional learning, (2) enhancing AI literacy through experiential learning with AI, and (3) designing an instructional model collaboratively informed by online teacher’s real-life practices, integrating a virtual human as pedagogical support. The intervention is currently in the design phase and, developed by the researcher with input from online teachers, using real cases from their daily practice to ensure relevance and authenticity.  Grounded in design-based research, it follows the five iterative phases of the ADDIE model: Analyse, Design, Develop, Implement, and Evaluate. The intervention consists of a four-week online module that integrates synchronous and asynchronous components, blending human and AI facilitation to foster creativity, engagement, and ethical awareness. Future phases will empirically examine online teachers’ experiences of learning ‘with’ guidance ‘from’ a virtual human and evaluate how this human-AI interaction shapes their understanding of AI’s pedagogical potential, where online teachers learn about AI while simultaneously learning with AI. Anticipated areas of investigation include transparency, emotional authenticity, and the distribution of pedagogical responsibility in human-AI collaboration. By combining instructional design, AI literacy, and professional learning theory, this work contributes to the development of human-centred, ethically grounded AI applications in online teacher education. It also seeks to advance the dialogue on how virtual humans can enrich networked online professional learning communities and support online teachers in shaping the future of education in the age of AI.