When AI Must Stop: NYU Tandon's Assignment-Level Framework

The key question is shifting from whether students may use AI to which parts of an assignment must remain the student’s work. NYU Tandon’s proposed framework classifies assignments as red, yellow, or green according to their learning objectives, with AI-free checks such as oral questions, in-class exams, and proctored quizzes when independent mastery matters.

The episode examines how Tandon developed the playbook through faculty review, departmental planning, and student policy votes. NYU reports that about 80 percent of full-time faculty participated in the review. Its November 12 announcement reports about 225 students in two Responsible AI courses debating and voting on course policies. The process demonstrates institutional participation and implementation planning, not yet improved grades, retention, mastery, or workload.

Research from Southern Cross University and the University of Dundee provides cautious supporting context. Their findings suggest assessment design and structured guidance may shift AI use toward verification, reflection, and supported learning rather than task completion. However, neither study proves that Tandon’s framework—or any particular policy model—causes better learning outcomes.

TOPICS:
- Assignment-level generative AI policy
- Independent mastery and assessment validity
- AI-free checks and mixed assessment portfolios
- Faculty and student participation in AI policy
- Structured AI use and assessment redesign
- Evaluating AI-supported learning

SOURCES:
- NYU Tandon School of Engineering — https://engineering.nyu.edu/news/nyu-tandon-unveils-new-strategy-and-playbook-ai-classroom
- NYU Tandon School of Engineering — https://engineering.nyu.edu/ai
- Southern Cross University — https://www.scu.edu.au/news/2026/can-universities-still-prove-students-are-learning-when-ai-can-do-the-assignment/
- Education and Information Technologies — https://discovery.dundee.ac.uk/ws/portalfiles/portal/165140909/s10639-026-13946-w.pdf

AUTHOR:
Ralph Aldredge, PhD, PE

WEBSITE:
https://AIHigherEdToday.com

YOUTUBE URL:
https://www.youtube.com/watch?v=S942h1a6uqA

DISCLOSURE:
This episode uses AI-generated narration.

DISCLAIMER:
This podcast is provided for educational and informational purposes. Although sources and content are reviewed before publication, information may be incomplete, contain errors, or become outdated as developments evolve. Listeners should consult original sources and exercise independent judgment before relying on the information presented.

EPISODE CREATION DATE:
November 14, 2026

EPISODE TYPE:
Full

COPYRIGHT:
© 2026 Ralph Aldredge. Original script and commentary. All rights reserved.
When AI Must Stop: NYU Tandon's Assignment-Level Framework
Broadcast by