Denying Alleged AI Use, Student Sues Yale SOM Over Year-Long Suspension

An executive MBA student at Yale University’s School of Management has filed a federal lawsuit against the university, alleging that he was wrongfully accused of using AI on a final exam and suspended by the school.

The lawsuit, filed February 3 in Connecticut’s U.S. District Court, was filed anonymously under the pseudonym John Doe. It seeks to overturn the plaintiff’s one-year suspension and have his academic record cleared of any wrongdoing, including the removal of the exam’s failing grade on his transcript.

It also seeks “a judgment awarding Plaintiff damages in an amount to be determined at trial, without limitation, past and future economic losses, loss of educational and career opportunities, damages to reputation, and loss of future career prospects.”

TO BAN OR NOT TO BAN?

It’s not the first time a student has sued a university – or even a business school – over AI use in classroom settings. But it does raise some sticky questions: What happens when universities and business schools rush to integrate AI into their curriculums while punishing students for using the very same tools? Where is the line between using AI to enhance productivity, stir creativity, and find new ways to approach a problem and, well, cheating?

Two months after ChatGPT went viral in November 2022, an operations professor at University of Pennsylvania’s Wharton School fed the chatbot his final exam for his MBA operations course. It would have earned a ‘B’ or ‘B-’, Christian Terwiesch, the Andrew M. Heller professor of operations told P&Q at the time.

Terwiesch simultaneously said that he would not outright ban the large language model from his classroom.

“If what we’re solving for is engagement, we shouldn’t be scared about the test question. When testing for certification and customizing learning, we may want to ban ChatGPT,” he said at the time. “But for solving for engagement, we should find other ways of engaging the students. And there I think the technology does miracles.”

That was several models ago, and the large language model has only grown more powerful.

Generative AI has evolved well beyond the trendy B-school buzzword. Institutions around the world are now racing to integrate it into courses and entire curriculums. They’ve created AI majors, degrees, and certificates. They encourage students to learn to use all the new, flashy iterations of a variety of AI tools, but are cracking down on those deemed to use it inappropriately.

While many institutions have rewritten honor codes and policies to cover AI concerns, there is no standardized approach in business schools – not that there necessarily should be. Innovation springs from doing things differently. But, without consistency, where is the line between cheating and using AI to enhance learning? And, how are students supposed to recognize that line?

“There is the tactical stuff in the classroom and then there is the reinvention of education,” Wharton’s Ethan Mollick told P&Q this fall. We selected the AI expert as our 2024 Professor of the Year for his “dogged intellectual pursuit of one of the world’s most disruptive technologies.”

“In business education, we teach people to be generalists. They are not a KPMG student or a Goldman Sachs student. They learn that through informal apprenticeship systems. At Goldman, they do the work over and over again and learn how to write a deal memo,” says Mollick, who requires his MBAs to use AI in his classrooms.

“But AI can now do that … There is a crisis of cheating but tests still work. If people cheat they cheat. I am much more worried about the next step in learning: the specialized piece.”

THE YALE LAWSUIT

In the Yale lawsuit, the unnamed plaintiff is a French entrepreneur and investor who is living in Texas, according to the Yale Daily News. He has a BSE in Biomedical Engineering and BA in Political Science from Rice University. He enrolled in SOM’s 22-month executive MBA program in July 2023 and was expecting to graduate this spring. The lawsuit claims the plaintiff “was on track to be the student Marshall/Valedictorian for his (EMBA) class.”

Poets&Quants reached out to the Yale School of Management with several questions about the lawsuit. The school declined to answer even the general questions about their AI policies, citing ongoing litigation.

A teaching assistant in the plaintiff’s course, Sourcing and Managing Funds, flagged his final exam in June because his answers were deemed too “long and elaborate” with “near perfect punctuation and grammar and elaborate formatting,” according to the lawsuit. The course was set for open book but closed internet, so no use of AI tools, according to the course syllabus.

A professor then ran the students’ responses through GPTZero, an AI detection tool, and found that some were likely AI generated. The professor referred the case to Yale’s Honor Committee.

What followed was a flawed disciplinary process, the lawsuit alleges. The student claims that he was pressured by Yale administrators to falsely confess to AI use, with one dean allegedly suggesting he could face deportation. The student also alleges that the Honor Committee proceeded with hearings without giving him access to key evidence, among other charges.

The lawsuit claims that Yale’s own policies prohibit using tools like GPTZero due to their high false-positive rates, particularly for non-native English speakers.

On November 8, the student says he submitted to the Honor Committee an email correspondence with Scott Aaronson – a prominent Computer Science professor at the University of Texas and an author of OpenAI’s original AI detection tool – about the unreliability of GPTZero. The student also submitted academic papers written by former Yale University President Peter Salovey, SOM Dean Charles Kerwin, and other Yale scholars that he had the tool evaluate.

“Those scans indicated that there was a ‘100% probability’ that selections from these works – some published over 30 years ago – were written by AI, a palpably false result,” the lawsuit says.

Later that day, the committee asked the student to return to campus, which he had already left, with his personal laptop. The student responded that he was not available. He was later informed that the committee found him liable for “not being forthcoming” to the committee. Yale gave him an “F” in the course and suspended him for a year. He appealed the committee’s findings but was unsuccessful.

In a response dated February 21, Yale has objected to the student’s request to remain anonymous. Yale lawyers argue that an independent assessment of the allegations requires him to be named. They also claim that details provided in the lawsuit are enough to identify the student and that he was previously accused of misappropriating non-profit funds, according to the Yale News Daily.

THE AI LAWSUITS HAVE BEGUN

The Yale case is just one in a string of recent student-inspired lawsuits in response to accusations of AI cheating.

Haishan Yang is suing the University of Minnesota after he was expelled for allegedly using AI on a preliminary exam last summer. Yang, who is not a native English speaker, was working on his second PhD from the school, and is seeking $575,000 in damages as well as $760,000 in defamation.

In Massachusetts, a family sued their child’s school district this fall after his high school alleged he used AI to complete a social studies project, causing him to earn a low grade. In the lawsuit, the parents argue that the school’s AI policy was vague and inconsistently enforced. The lawsuit contends the action hurts the student’s chances of getting into Stanford University and other top schools.

And then there’s the case of Eightball, an AI-powered study tool designed to help students generate study materials based on uploaded course content. It was co-founded by three undergrads at Emory University, including a business major at Goizueta Business School. The tool won Emory’s annual 2023 Pitch the Summit startup competition, earning the $10,000 grand prize, and was featured on Goizueta’s website. (The post has since been removed but is archived here.)

Months later, Emory suspended two of the students, accusing them of violating the university’s honor code. The university said the tool could be used for cheating even though no evidence was produced to show that students were misusing the tool, according to his lawsuit.

One of the co-founders, Benjamin Craver, sued Emory in May, according to and article in the Business Insider. The student argues that the university unfairly punished him while simultaneously celebrating AI innovation.

Through a spokesperson, Emory University declined to answer Poets&Quants questions about the lawsuit.

You can read the full Yale lawsuit, and Yale’s first response, in this article.

DON’T MISS: CHATGPT: HOW TO USE IT IN A BUSINESS SCHOOL CLASSROOM AND SO HOW DID CHATGPT FARE IN MBA ESSAYS?