Artificial Intelligence is rapidly becoming embedded in teaching, learning, and assessment—yet much of today’s educational AI relies on opaque, proprietary systems, raising urgent questions about transparency, equity, data control, sustainability, and academic freedom.

This workshop introduces the PublicAI movement as a compelling alternative aligned with the values of open education, open pedagogy, and collective knowledge creation. Using concrete initiatives—including open foundation models, shared datasets, and public inference infrastructure—we will show how PublicAI can support teaching, research, and institutional innovation without reinforcing monopolization or black-box decision-making.

A special focus will be on Apertus, the first fully open-source Swiss AI model, illustrating how national and academic communities can reclaim agency in the AI landscape.

In the hands-on part, participants will use the PublicAI API to build a Personalized Q&A Generator that transforms lecture materials into an adaptive learning experience. Instead of static quizzes, the AI-driven engine:

• Generates targeted questions directly from course materials or selected sections on demand
• Adapts in real time, modifying difficulty and focus based on student responses
• Supports formative assessment and self-directed learning

Through guided steps and accessible technical explanations, participants will gain practical experience building an AI-powered tutor that bridges teaching content and student mastery, while remaining transparent, auditable, and aligned with open education principles.

Note: The first part is suitable for both technical and non-technical participants, while the hands-on workshop is geared toward those with some technical experience.