Class-wide Policy Memo Assignment

Overview

In this assignment, you'll work together across the sections of “Citizenship in the 21st Century” to formulate a 350-word policy memo that attempts to represent the views of the first-year class about one of our most pressing questions: How should the Stanford community promote learning, reduce cheating, and conduct accurate assessments of student learning in the age of artificial intelligence?

You’ll draft this memo in a small group of four or five students, get feedback from your section and revise, and then read and rank the memos written by other sections. The authors of some of the highest scoring memos will be invited to a lunch to pitch and discuss their proposal with Stanford’s provost (the senior academic officer), the vice provost of undergraduate education (VPUE), and university experts on AI and education. We can’t promise that your memo will be adopted as policy, but we think you’ll benefit from this experience of citizenship as participation in a self-governing group, using persuasive writing to try to bring about a class-wide consensus on an important issue—an issue where you may have direct knowledge, broad support will be required for any new policy to succeed, and nobody knows the right answers.

COLLEGE102 instructors Prof. Keith Winstein and Prof. Emilee Chapman provide a video introduction to this assignment here.

Background 

Stanford has a long history of dealing with these questions in a student-led way. In 1921, the students successfully petitioned the faculty that they, not the faculty or the university administration, should be granted the power to govern themselves on matters of academic honesty. In the “Honor System” of 1921, Stanford’s students asked to take over the responsibility of educating new students about academic honesty, deterring cheating among themselves, and when they learned of infractions, reporting them to “student councils” which would decide on the appropriate consequence for violators. In exchange, the faculty agreed not to police students, proctor exams, or create assessments that tempted students to cheat, and agreed to report violations to the student councils if they discovered them.

This system became the “Honor Code,” and it stood with minor revisions for a century, at least on paper. As early as 1930, student leaders complained that students weren’t holding up their end of the bargain and the Honor Code system risked collapse. Students pledged renewed efforts, but over the subsequent decades, it became extraordinarily rare for students to report academic dishonesty; from 2018-2021, only 2 of the 720 Honor Code cases were filed by students. Some faculty were concerned that they were prohibited from proctoring exams because students had promised to discourage and police one another’s cheating but were not doing so. Some faculty lost confidence that the grades they assigned were accurate summaries of how well students had accomplished a class’s learning goals, or fair to the students who accomplished learning goals without cheating. Meanwhile, most students no longer viewed the Honor Code as an exercise of student self-governance and power; it was more commonly perceived as a punitive system of the university administration or faculty. Some students and alumni viewed the “consequences” process as secretive, slow, and unfair, and began to describe violations of the Honor Code as something that happened to a student (“he got Honor Coded”) rather than something a student did. Even if the consequences were decided partly by fellow students, most students viewed “snitching” on violators as morally wrong and socially anathema. Students complained that there was no way to report faculty for violating the Honor Code, for example, faculty who published unfair assignments or unclear guidance on permissible collaboration in their classes.

From 2019-2023, student-faculty committees (with equal student and faculty membership) debated revisions to the Honor Code, and ultimately proposed to remove the notion of the Honor Code as a student-led undertaking, to remove the requirement for students to deter or report violations by fellow students, to reduce the penalties (removing any disciplinary notation for students who don’t contest a first-time infraction), and to charter a new Academic Integrity Working Group to study the causes of academic dishonesty at Stanford and conduct a trial of in-person proctoring of exams, which (depending on the results) could lead to the permanent introduction of exam proctoring. To take effect, this package of reforms had to be approved by the Undergraduate Senate, the Graduate Student Council, the Faculty Senate, the Board of Judicial Affairs, and the university president; each approved it in 2023, but only after the Faculty Senate unilaterally adopted the Honor Code revisions and proposed to begin proctoring exams that fall unless the student governments approved the full package.

Where things stand now

Three years later, the Academic Integrity Working Group is ongoing, and many professors and students believe the world has changed dramatically. While Stanford researchers have been working to create AI since the early 1960s, the recent rise of commercial chatbot services (especially ChatGPT in 2022) has affected many sectors of society, including education. Most academics believe these services have led to more cheating, and perhaps (but not as clearly) less learning. Earlier this year, New York Magazine claimed in a widely read article that “Everyone’s Cheating Their Way Through College.” More nuanced accounts of student use of AI appeared in The Chronicle of Higher Education and Inside Higher Ed. Instructors from across departments have grappled with the appropriate response now that ChatGPT can produce passable work in most college assignments, from computer science to classics.

In 2025, many students report wrestling with the anxiety of wanting good grades and other tangible indicators of success to ensure the best chance of a good job after college--indicators more tangible than the academic learning and internal transformation of a college education. Some students view parts of the college curriculum as gatekeeping exercises; they aren’t sure that every class is offering knowledge that will prove valuable in the future and is worth the time and effort, especially if ChatGPT can do the work. Some faculty fear that many students aren’t learning as much as they used to before AI capabilities advanced. Many faculty at Stanford view their goal as fostering student learning in a beloved intellectual area and see assessment and grading as a smaller part of the job; these faculty resent being asked to “police” students.

Nobody has the answers about what to do. Some faculty think the only way to stay sane is to focus on the students who want to learn the material, ignore the ones who don’t, and avoid expending effort trying to police cheating. Others believe this is unfair to students who accomplish learning goals without cheating, and think the best solution is to shift grading away from homeworks (where it’s difficult to assess how well students have learned given the prevalence of AI use) and towards proctored, in-person written or oral exams where cell phones and Internet-connected laptops would be prohibited. Some other faculty think we should give up on trying to measure learning precisely with letter grades, and shift to a less fine-grained system that acknowledges the new uncertainties of measuring student learning (for example, more classes would be credit/no credit). Some faculty are working to create new curricula that focus on helping students learn knowledge and skills they believe will be needed even with a pervasive societal reliance on AI chatbots--but much of the material likely to fall in those categories (critical thinking and reasoning, grounding in an intellectual discipline, and a liberal-arts education more generally) are already longstanding goals of a Stanford education.

Your assignment

The question in this assignment is: what do you, the first-year class, think our university community should do to promote learning, reduce cheating, and conduct accurate assessments of student learning in the age of artificial intelligence? The product should contain your ideas, persuasively justified, and written in a way that commands support from other sections of “Citizenship in the 21st Century.” The goal is to practice intellectual leadership and collaborative writing, engage in meaningful peer review, build a consensus around your ideas, and participate in real citizenship that can change university policy. Your group could write a radical proposal, or a modest one, but whatever it is needs to try to accumulate support from a broad swath of your class. Political Scientist Danielle Allen, whom you’ll be reading later this quarter, shows in Our Declaration that people can use writing -- writing “memos” like this one -- to build consensus and legitimacy to create new forms of organization. As she says, “those who write the best memos set policy for businesses, cultural organizations, and governments. … [Writers] wield the instrument by which our world is organized” (73). You’ll use writing to help shape Stanford’s future and that of higher education and our society.

Working in groups of 4 or 5 (four groups per section), craft a policy memo of no more than 350 words that proposes 1-2 concrete actions or policies that Stanford students, faculty, and/or administrators should take or adopt regarding the above topics. Your proposal should be clear, persuasive, and executable—something students, faculty, or university officials could actually implement. What you propose and what you read and cite to support your proposal is entirely up to your group. We’d suggest that you start by reviewing the materials linked above, including the Stanford Honor Code and the charter of the Academic Integrity Working Group.

Memo-writing tips

Make sure the concrete actions or policies are clear at the start, so the reader knows what you are proposing. Then, give the 2-3 best reasons why these actions should be taken. Marshal facts with citations in support. Give the best arguments for why students and faculty should both subscribe to your ideas (realistically, any policy will need both student and faculty buy-in to happen). To build persuasiveness and credibility, include at least the strongest reason you can think of for “here’s why this might be a bad idea”--then argue against that objection. (It does you no good if the reader can think of a different, stronger, reason why your proposal is lousy--the memo will lose force if it doesn’t anticipate and respond to their strongest objection.) Keep the word count at 350 words maximum.

After drafting your memo, you'll get feedback from others in your section, and (optionally) combine memos to form the strongest consensus output from your section. Then, you’ll participate in a cross-section review process where you'll read and rank eight memos from other sections. Your ranking will carry higher weight if your entire section can agree on them. The authors of the highest ranked memos will be invited to a lunch to pitch and discuss their proposal with Stanford’s provost (the senior academic officer), the vice provost of undergraduate education (VPUE), and university experts on AI and education.

Timeline and Process

Week 1

Form your group and start brainstorming. Practice active listening as you hear different perspectives from your teammates: How are LLM tools such as ChatGPT a threat–or a boon–to teaching, learning, and assessment at Stanford University? What can we do to respond?

Week 3

• Paste your 350-word draft memo into the Canvas discussion by Sunday, January 18, 11:59 pm
• Read the other three memos from your section.
• Session 2: Spend 30 minutes in class debating the merits of the draft memos with other groups. Which will you push forward?

Weeks 4 and 5 

Continue asynchronous / out of class revision in a google doc. (Be sure these links are posted to this assignment’s Canvas discussion so that your section can continue the conversation!) By the end of week 5, your instructor will then submit your section’s revised memos to the “rotisserie,” a google form that will help us randomly distribute each section’s memos to other sections.

Weeks 6 and 7

Read eight memos from other sections (these will be distributed by your instructor over Canvas announcement); decide which are your top two and why.

Week 7.2

Arrive to class with your top two memos in mind. In 30 minutes in class, you’ll attempt to come to a section-wide agreement on which two memos of the eight could represent the whole first-year class. This is where you'll experience firsthand how challenging collective decision-making can be when people disagree.

Week 9

Authors of the highest-ranked proposals meet with university officials to present and discuss their ideas.

Grade

Completing a thoughtful 350-word-max memo draft in Week 3 with a small group of your classmates – and responding to the other three draft memos in your section – will earn you 5 points. Contributing to the revision of class memos and ranking others’ memos will also contribute to your participation grade. You'll also have the option to reflect on this collaborative experience in your final essay this quarter, analyzing what you learned about productive disagreement and group decision-making.