Brandeis partnering on major new National Science Foundation grant to build an AI-literate workforce
Brandeis is part of the NSF National AI Institute for Student-AI Teaming (NSF iSAT), which has secured a five-year, $20 million renewal. The institute is led by the University of Colorado Boulder with collaborators at Boston University, Colorado State University, the University of California Berkeley, and Worcester Polytechnic Institute. James Pustejovsky, the TJX Feldberg Professor of Computer Science and of the Volen National Center for Complex Systems at Brandeis, serves as a co-principal investigator.
What's been built since 2020
NSF iSAT launched in 2020 to make classrooms more effective and engaging. The team developed two AI partners that help small student groups learn together by steering discussion, exploration, and reasoning in coordination with teachers. More than 6,000 middle-school students and educators have used these tools and new AI curricula.
- AI partners facilitate group talk, productive disagreement, and evidence-based reasoning.
- Teachers stay in control; the AI augments classroom flow and feedback loops.
- Focus is on small-group dynamics where most peer learning happens.
Next phase: AI literacy and classroom teaming
The renewed grant targets an urgent national need: an AI-ready workforce. NSF iSAT will continue advancing AI support for small-group learning and will co-develop a semester-long AI literacy curriculum to help students work capably with AI systems.
Brandeis contribution: human signals, better teaming
Pustejovsky's lab studies multimodal communication-how speech, gesture, gaze, posture, and action interact during learning and problem solving. Those signals feed dynamic models of students' knowledge and shared context, enabling AI to reason about group interaction in human-centered ways.
As Pustejovsky explains, this line of research builds AI partners that support collaboration, sense-making, and wide participation in classrooms. These foundations matter for education and for preparing students to work productively with AI across future roles.
Why this matters for science and research professionals
- Methodological depth: multimodal modeling connects discourse, embodiment, and task progress-useful for HCI, learning analytics, and collaborative systems research.
- Ecological validity: findings come from classroom deployments with thousands of real learners and teachers.
- Human-centered AI: emphasis on equity, participation, and explainable interaction within group workflows.
- Workforce alignment: students practice teaming with AI-skills that translate to labs, fieldwork, and applied R&D.
Learn more and explore training
For program context, see the NSF National AI Research Institutes and the Institute for Student-AI Teaming.
If you're building AI literacy pathways for your team or students, you can browse structured options by role on Complete AI Training.
Your membership also unlocks:
