Cornell University is launching two new AI in education initiatives backed by a $2 million gift from the Dake family. The programs, announced September 17, 2026, will fund college-specific experiments in teaching and expand a student research fellows program - both designed to test where AI helps learning and where it doesn't.
College-specific pilots will test AI's fit across disciplines
The Cornell AI Initiative's new AI Integration Pilots give individual colleges room to experiment on their own terms. Deans have been asked to identify a core teaching element - a skill, course sequence, or subject area - that could be reformed through AI. Each college will assemble a project team of three to five faculty members, including a faculty lead. The teams can propose anything from adopting AI tools that build student skills to redesigning assessments so students can't lean on AI in ways that short-circuit learning.
"Teaching unfolds differently, learning unfolds differently, in different academic areas, and that's not going to change," said Steven Jackson, vice provost for academic innovation. "If anything, AI may, in fact, make it more distinct." Jackson said Cornell is one of very few institutions taking such a deliberate approach focused on "the nuance of using, and not using, AI in instruction and learning."
Teams will implement changes across spring 2027, fall 2027, or spring 2028 semesters. A graduate student or postdoctoral researcher will be embedded in each team for two years to handle assessment and coordination. The Center for Teaching Innovation and the AI Innovations Lab will provide pedagogical and technical support. Results will be shared across the university in 2028.
Undergraduate fellows act as classroom researchers
The second initiative expands the AI Undergraduate Fellows program. Students serve as embedded researchers in their own classrooms, working alongside faculty to identify where AI might fit, test those methods, and evaluate what happens. Their work also feeds into broader Cornell research on AI in teaching, tracking learning outcomes and student experience.
"Students have insights into the student experience and student needs that faculty don't have, that staff don't have," Jackson said. "They can shape our thinking, and they're playing a pretty important role in research." The program reflects what Jackson called the initiative's "students-as-partners" philosophy - treating AI as something done with students, not to them or for them.
Why this matters for educators
Cornell's structure offers a template for how institutions can move past blanket AI policies - either banning tools or adopting them wholesale - and instead make field-specific decisions. The pilots acknowledge that an appropriate use of AI in a computer science course may look nothing like its use in a creative writing workshop. For K-12 teachers and administrators watching higher ed for signals, the takeaway is clear: the most serious AI for Education efforts are now asking not just whether to use AI, but where it belongs in a specific subject's learning sequence. Programs like the AI for Teachers learning path can help educators build the fluency needed to make those judgment calls in their own classrooms.
Your membership also unlocks:
