Princeton Ends 133-Year-Old Tradition as Data Shows AI Cheating Cases Increase

Princeton’s Honor Code

The Honor Code at Princeton has been a tradition at the prestigious university since 1893. Under this system, professors trusted students to complete examinations without supervision, and students pledged not to cheat and to report any suspected violations of this code.

According to Princeton’s Honor Committee:

The system has been viewed as a point of pride for the university for 133 years. However, this is now changing due to concerns regarding AI.

According to The Wall Street Journal, the faculty voted on May 11 to require proctoring in all in-person exams from summer after significant numbers of undergraduate students and faculty requested the change.

Data Demonstrates Growing AI Concerns

Princeton’s Honor Committee publishes statistics regarding alleged academic integrity violations each year. These reports demonstrate that unauthorized use of AI has grown significantly from 2015 to 2025.

The Fall 2015–Spring 2020 report does not include any language about AI. However, the most recent report (Fall 2020–Spring 2025) explicitly indicates that “generative AI” falls under the category of “Unauthorized use of outside materials.”

A comparison between the two reporting periods also reveals an increase in cases requiring further investigation and disciplinary action.

Source: https://honor.princeton.edu/statistic-reports

While the number of initial reports remained the same, the number of reports that appeared legitimate enough to move to a hearing nearly doubled. This indicates that a much larger percentage of reports warranted a hearing and further investigation—from 39% to 73%. Subsequently, the number of cases that the committee found responsible for violating the code and warranting disciplinary action more than doubled.

In addition, it is likely that many more cases of cheating go unreported as students are increasingly reluctant to report their peers.

According to the 2025 survey of seniors by the student newspaper, The Daily Princetonian, the number of seniors in 2025 who admitted to using AI for an assignment when it was not allowed rose to 27.7%. In addition, 44.6% of those surveyed indicated that they had knowledge of a peer violating the Honor Code and chose not to report it.

It is therefore no surprise that faculty and students alike requested and supported this change.

Princeton’s decision and its students’ attitudes follow a broader trend in universities with similar longstanding policies. Stanford began piloting the proctoring of in-person exams in 2023 despite student opposition. When the university voted to allow the practice in April of 2026, the committee noted that more students seemed to be supportive.

As AI tools grow increasingly effective and accessible, we will likely see more institutions reevaluate their policies and make amendments to maintain academic integrity and high academic standards.