Science teachers are experimenting with AI, but many find themselves stuck at the surface level of classroom use. Our survey of ten participating middle and high school teachers confirmed this. The group was already exploring AI broadly, collectively naming over 15 different tools, and every participant reported modifying AI outputs before use. Still, they expressed a clear desire to move beyond generic content generation toward integrating AI in content-specific ways, grounded in scientific practices like argumentation and modeling.
This is exactly what the National GENIUS Center set out to do in a four-hour session on March 28, 2026, on the University of Georgia campus. Starting from a classroom case study that contrasted meaningful and superficial AI integration, teachers worked their way toward designing their own argumentation-worthy tasks. Along the way, they explored the GENIUS platform’s ArguAgent and ModelAgent hands-on and practiced using the RISEN prompt framework to craft lessons aligned to the Georgia Standards of Excellence.
Post-session surveys showed that the experience landed: teachers reported greater confidence in using AI for science instruction and rated the hands-on activities and facilitator support highly.
The PD is part of our broader research effort to identify how professional development can best support science teachers’ AI literacy and classroom AI integration. Based on our findings and teacher feedback, the PD design will continue to evolve.
We look forward to many more opportunities to work with teachers. This group taught us a great deal, and we are grateful for their time and insight!
