Starting next fall, every first-year student at the University of Chicago will take a writing course where AI is not allowed. At Eastern Illinois University, students are asking for the opposite kind of rule: limits on how their professors use AI. And at NYU Tandon, faculty released a guide that sorts assignments into red, yellow, and green depending on whether AI may be used at all. The stories from November 8 to 14 point the same way. Institutions are moving past blanket permission and blanket bans toward a harder question: what must a student be able to show without AI, and how will anyone know?
“AI gets in the way, because it will automatically do all the things that you should be learning yourself.”
Sara Sood, Associate Dean for Academic Programs, McCormick School of Engineering, Northwestern University (Holland & Pomer, 2026)
NYU Tandon’s “Protecting Learning” guide sorts assignments into red, yellow, and green
The Details
On November 12, NYU Tandon School of Engineering released Protecting Learning, a practical guide to AI-use policy for instructors, departments, and schools. It grew out of departmental assessments that fed working groups on policy, student use, and teaching resources under Tandon’s AI Across the Curriculum initiative, led by Julia Stoyanovich. Red assignments prohibit AI, yellow assignments permit limited assistance, and green assignments make AI part of the assessed work. Instructors set a course baseline and can change the category assignment by assignment. Where independent competence matters, the guide recommends checks completed without AI, such as oral questions, supervised quizzes, or classroom exams. It also states that an AI-detector score alone cannot establish a violation; instructors should gather multiple forms of evidence and speak with the student before reaching a finding. The guide includes templates for involving students in setting course rules, with the instructor keeping the final decision. It is not official NYU Tandon policy. Public comment is open through November, with a first revision planned for January 2027 (CASRAI, 2026).
Why it Matters
The guide gives faculty a workable middle path between “anything goes” and “never.” An early assignment that builds foundational knowledge can be red; a later one that tests whether students can use AI well, verify it, and defend their decisions can be green. The point is that the choice is deliberate and stated in advance. Its stance on detectors matters too: a score is a reason to ask questions, not a verdict. The guide has not yet been evaluated for its effect on learning, but it offers a ready template for course and department policy.
Employers say new graduates struggle most with verifying AI output and AI ethics
The Details
University of Alabama researchers Laura McNeill, Mohammad Mohi Uddin, Stephen Abu, Mengshi Pei, Idowu David Awoyemi, and Lesley Regalado published a four-phase mixed-methods study on November 12. They reviewed 56 sources, held focus groups and interviews with 13 industry leaders, built a survey from those themes, and surveyed a convenience sample of 110 mostly U.S. employers. Employers described AI readiness as both technical and human: soft skills and human oversight were rated as highly as technical AI skills, and human oversight had the highest mean importance. They rated recent graduates as underprepared, with the lowest scores for ethics and policy and for verifying AI outputs. The authors recommend five undergraduate competency areas: AI literacy, prompting, and tool fluency; critical evaluation and output verification; ethical, legal, and responsible use; human-AI collaboration and soft skills; and hands-on application. The sample is a convenience sample, the survey instrument is not validated, and a link between governance and efficiency is exploratory, not causal (McNeill et al., 2026).
Why it Matters
A graduate who can produce a report with AI but cannot tell whether it is accurate is not ready for the job. For business programs especially, students should be able to spot unsupported assumptions in AI-generated analysis, test recommendations against credible evidence, and explain why a decision is sound. Those abilities depend on disciplinary knowledge and judgment, which means AI literacy belongs inside the major, assessed through real tasks, rather than in a single stand-alone course.
Policy & Governance
-
A proposed model for assessing what students can do with and without AI
Samar A. Ahmed of Rabdan Academy (UAE) and Ain Shams University (Egypt) and Mohamed H. Shehata of Arabian Gulf University (Bahrain) and Capital University (Egypt) propose AI-Integrated Programmatic Assessment in a conceptual analysis published November 13. The model gathers three streams of evidence on the same competency across many assessments: performance without AI, performance with AI available, and the quality of a student’s judgment, verification, and responsibility when using AI. A gap between assisted and unaided performance is treated as a question to investigate, not proof of harm; the authors describe cognitive erosion as “a testable possibility rather than an established general effect” and note the model itself has not yet been tested (Ahmed & Shehata, 2026). For program leaders, it offers a structure for collecting evidence of both independent mastery and responsible AI use, rather than relying on any single assignment.
-
Northwestern leaves AI rules to schools and instructors, within clear limits
At Northwestern, AI policy is set mostly at the school level, with instructors deciding the details. In the Weinberg College of Arts and Sciences, students may not present AI-generated work as their own, instructors must give clear guidelines, and faculty and graduate students may not use AI to grade undergraduate work. Medill students need explicit instructor permission and must explain how they used AI. Faculty approaches range widely: a history course with a strict no-AI policy, a French program that bars beginners from AI and allows advanced learners to use it with permission, computer science homework where AI is allowed but students must be able to explain their answers, and an introductory programming course that bans AI until students learn the fundamentals (Holland & Pomer, 2026). Decentralized policy can fit disciplinary needs, as long as every student knows the rule before the work begins.
Programs, Research & Infrastructure
-
AI-supported research training for future teachers, with clear limits on what it shows
Aygerim Boranbay, Gulnara Karbozova, Zhanna Makhanova, and Farida Salybekova studied 133 second-year preservice English teachers at Zhanibekov South Kazakhstan Pedagogical University, in a study published November 12. Over eight weeks, 67 students used ChatGPT in a sequence that required them to attempt each research task on their own first, explore AI suggestions, verify and critique them, revise, and make the final decision themselves; 66 students received conventional instruction. The AI cohort scored 1.34 points higher on a 36-point research competence scale after adjusting for baseline, with gains in methodological and cognitive competence and no significant difference in digital or ethical competence. Each condition was a single intact group, so the authors state the results describe cohort differences and cannot establish that AI caused them (Boranbay et al., 2026). The instructional sequence, though, is a useful model: students do the thinking first, then use AI to test and refine it.
-
Harvard launches an AI guidance tool built on 5,000+ studies
On November 10, the Harvard Graduate School of Education released the AI Guidance by Age Matrix through its new Compass Collaborative, led by Jeff Curley and Tessa Forshaw. Drawing on more than 5,000 research pieces across 81 developmental areas, the tool lets educators, parents, students, and policymakers choose a role, a developmental domain, and an age range from early childhood to adulthood to see where AI may help or hinder growth. Its central finding is that AI’s effects depend on the learner’s age and the skill being developed; answer tools, for example, may undercut the retrieval practice that builds memory. Forshaw’s rule of thumb: uses that move toward skill practice or human relationships are generally promising, and those that move away deserve a second look. One assistant principal said the tool “still feels pretty basic” (Langreo, 2026). Its coverage of college-age learners is not described in detail, but the principle carries over directly to course design.
Other
-
University of Cincinnati approves a humanities and technology center
UC’s Board of Trustees voted in October to establish the Center for Humanities and Technology, announced November 12. Funded by a $500,000 National Endowment for the Humanities award and led by philosophy professor André Curtis-Trudel, it will support humanities research on AI, host a speaker series and an annual conference, and run AI ethics workshops for high school students with the Cincinnati Ethics Center (University of Cincinnati, 2026).
-
Two convenings bring AI and education practitioners together
Kansas State University hosted its hybrid Heartland AI Symposium, “Our Society in the Age of AI,” from November 10 to 12, with tracks including AI literacy, AI ethics, education’s and students’ responses to AI, and AI and writing, plus workshops and community town halls (Kansas State University, n.d.). American Military University and American Public University held their third annual AI and the Future of Education Conference online November 12 to 13, covering teaching, student support, accessibility, equity, and ethics for research review boards (American Public University System, 2026). Convenings set agendas; their value shows in what participants change afterward.
Do It Now Checklist
Betting On: Demonstrated Learning
This week’s stories put the weight on evidence of what students can actually do. Chicago is protecting foundational writing. NYU Tandon gives faculty a way to decide, assignment by assignment, where AI belongs and how independent competence will be checked. Employers say graduates fall shortest in verifying AI output and judging its ethics. The institutions that align assessment with disciplinary knowledge and professional responsibility will have real grounds for saying their graduates are ready.
With Inspiration Moments, we share motivational nuggets to empower you to make meaningful choices for a more fulfilling future. Remember: “Life happens for you, not to you, to live your purpose.”
Respectfully,
Dr. Lynn Austin
