Generative AI arrived in higher education faster than anyone's policy for it. Instructors are improvising — writing course policies, revising assignments, renegotiating what their expertise is for — and the futures they imagine while doing so shape what they build now.
This project studies that improvisation directly: how programming and writing instructors are actually forming policies and adapting pedagogy, what AI social presence does to perceived learning and satisfaction, and where an AI instructor falls into the uncanny valley instead of helping.
It also holds the lab's applied work on educational AI agents: teaching signals drawn from homework, bounded confidentiality when a learning signal becomes a safety signal, and governance for multilingual tutors that flatten the cultures they translate between.
Themes across projects
- Disclosure, trust, and accountability in human–AI teamsAlso in AI-Mediated Communication
- AI-mediated learning and communicationAlso in Digital Literacy at Scale
- Methodological pluralism — questions shape methods, not dogmaShared by every project
Publications
Published
- 2025
Unpacking the dilemma: the dual impact of AI instructors’ social presence on learners’ perceived learning and satisfaction
Chen & DiFranzo · ACM Web Science 2025, 22–31
Workshops & extended abstracts
- 2026
ClassPulse: a bidirectional AI-supported programming homework platform for teaching signals and student learning support
Shi & DiFranzo · AIED 2026 Late-Breaking Work, 282–288
- 2026
The fluency trap: culture-aware governance for multilingual LLM tutors
Shi & DiFranzo · CATS 2026 Workshop @ AIED
- 2026
Strategic missingness in AI help-seeking traces for programming education
Shi & DiFranzo · CSEDM 2026 Workshop @ AIED
- 2026
When learning signals become safety signals: a bounded-confidentiality framework for educational AI agents
Shi & DiFranzo · IRAISE 2026 Workshop, PMLR, 123–128
Preprints & work in progress
- 2026
Designing and evaluating AIs for asynchronous online learning
Chen & DiFranzo · under review · CSCW 2027
-
Understanding programming and writing instructors’ strategies for responding to generative AI
Steup, Wegrzyn, DiFranzo & Baumer · in preparation · target CHI