---
title: "Dartmouth’s Provost Wrote About AI Cheating and the Detector Had Questions"
description: "An op-ed on AI, a defense of the op-ed, and a statement defending the defense, all allegedly written by AI"
author: "Nate Ledger"
published: 2026-09-25T14:51:57.001Z
modified: 2026-09-26T17:25:25Z
url: https://rews.cc/a/dartmouth-s-provost-wrote-about-ai-cheating-and-the-detector-48db18
language: en
tags: ["ai", "education", "ethics", "dartmouth", "openai", "us"]
publisher: "Rews (https://rews.cc)"
---

# Dartmouth’s Provost Wrote About AI Cheating and the Detector Had Questions

*An op-ed on AI, a defense of the op-ed, and a statement defending the defense, all allegedly written by AI*

By Nate Ledger · September 25, 2026 · https://rews.cc/a/dartmouth-s-provost-wrote-about-ai-cheating-and-the-detector-48db18

## In brief

- Dartmouth provost Santiago Schnell’s August Washington Post op-ed on AI cheating was flagged as 100% AI-written by Pangram, per a Semafor analysis
- The Dartmouth found a median 96% AI-written score across nine Schnell works from 2026; his pre-ChatGPT writing tested 100% human
- Schnell’s emailed defense to the student paper also tested 100% AI-written; he calls AI an “assistive” tool and says he owns the content
- Schnell apologized for not disclosing AI use; a later letter to the editor tested 100% human-written
- President Sian Beilock announced a review of Schnell’s AI use; an AAUP petition seeks his removal

Here is a hypothetical that is not actually hypothetical. Imagine you are the chief academic officer of an Ivy League college — the person ultimately responsible for academic integrity. You publish an op-ed in a major national newspaper arguing that universities are fighting AI cheating but missing a deeper problem. An AI-detection firm then announces that your essay about AI was, by its measure, 100 percent written by AI. The student newspaper investigates further, finds a median “AI-written” score of 96 percent across nine pieces you published this year, and asks you for comment. You email a statement defending yourself. The students run the statement through the detector. It comes back 100 percent AI-written.

This is the situation now consuming Dartmouth College, where Santiago Schnell has been provost for a little over a year. In August he wrote an op-ed for the Washington Post titled “Universities are fighting AI cheating. But there’s a deeper problem.” There was indeed a deeper problem, just not the one in the piece: on Aug. 26, the news platform Semafor published an analysis of AI writing in national publications in which the detection tool Pangram identified Schnell’s essay as 100 percent AI-written. Schnell’s essay urged universities to teach students critical thinking alongside intelligent AI use.

Schnell told Semafor the essay was based on his own thinking and that he used ChatGPT “to refine a few arguments, check for grammatical errors, and submit it,” followed by multiple rounds of editing. A native of Venezuela, he also noted there is evidence AI detectors “are biased against non-native English speakers, such as myself.” Fair enough; that is a real documented concern about these tools, and for a while the story seemed to fade.

Then, on Monday, reporters Alex Klee and Isabela Pierry at The Dartmouth published their own analysis of Schnell’s opinion pieces and academic articles from 2026. The clever part is that Schnell, a professor of mathematics who has served on the editorial boards of ten leading scientific journals, brought his own control group: Pangram 4.0 labeled all of his written work from before ChatGPT’s November 2022 debut “100% human-written,” while nine works from this year returned a median “AI-written” percentage of 96 percent. As experimental designs go, it is hard to beat “the same author, before and after.”

On Sept. 18, Schnell emailed a statement to the student paper explaining that he uses AI to refine his writing and become more efficient, while taking responsibility for the content. The students ran the statement through the detector. One hundred percent AI-written. Or, as one Reddit user summarized the situation: “Dartmouth Provost caught using AI for everything.”

> “I regard these technologies as assistive tools. They do not identify the problem, design the research project, guarantee the validity of research results, interpretations, citations, conclusions or substitute for my research, intellectual judgment or authorship.”

Which is a perfectly reasonable position for a provost to take about his own writing, and a somewhat stranger one for a chatbot to take about itself, which — if the detector is right — is what happened. There is a recursion problem here that no amount of refining resolves.

In a statement to the Boston Globe sent Thursday night, Schnell shifted to contrition: “In retrospect, I now realize I should always explicitly disclose use of AI in my public writing and be attentive to the expectations of AI use set by publications. I am sorry that this was not my standard practice and, moving forward, it will be.” On Friday, The Dartmouth published a letter to the editor from him in a similar vein. “I did personally run it through Pangram,” said Eli Moyse, the student columnist who has called for Schnell’s firing. “It came back as 100 percent human written, which I guess is a step in the right direction.” That sentence is doing a lot of work, and it knows it.

The awkward backdrop is that Dartmouth has spent this year celebrating 70 years as the birthplace of artificial intelligence — the 1956 Dartmouth Summer Research Project on Artificial Intelligence is widely considered a founding event of the field — and marketing itself accordingly. Last year it became the first Ivy League school to deploy AI at what it calls institutional scale, through partnerships with Anthropic and Amazon Web Services, with a faculty leadership group that, per the college, “ensures the technology serves Dartmouth’s educational mission, not the other way around.” Meanwhile Dartmouth’s own policy states that students “may not use GenAI tools for submitted coursework unless expressly permitted.”

So the provost’s position — the rules for me are assistive, the rules for you are prohibitive — has not gone over well. “It’s a textbook case of exactly what we tell our students not to do,” said religion professor and department chair Christopher MacEvitt, adding that his “jaw literally dropped” reading the student paper’s reporting. History professor Bethany Moreton asked what the college expects faculty to tell students when AI rules are “optional for people in power.” Ben Stevenson, a history major, called the situation “Orwellian” and the school “a national embarrassment.”

President Sian Beilock said Thursday that “the questions this moment raise go well beyond any one person” and that “it is more clear than ever that we need role-appropriate rules and policies for this technology,” then emailed campus Friday to announce “a rigorous, objective review of how \[Provost Schnell\] has used AI, measured against the criteria of the publications to which he submitted his work.” It caps what MacEvitt called a “terrible September” for Beilock, who is also freshly portrayed as a capitulator-in-chief in Jacob Weisberg’s new book, *Profiles in Cowardice: A Study of Collaboration in the Trump Era*. The Dartmouth chapter of the American Association of University Professors is circulating a petition for Schnell’s removal: “How do we ask our students and colleagues to follow the rules when the academic leader of Dartmouth is allowed to break them?” Alum Unai Montes-Irueste, writing in The Dartmouth, accused Beilock of selling “a pro-AI Dartmouth to prospective students and donors,” and told the Globe many in the community see “a tremendous hypocrisy.”

The genuinely interesting question, though, is the one Dorothy Leidner, a professor of business and AI ethics at the University of Virginia, raised in an email to the Globe: who is the author here? Researchers have long used graduate students to build the intellectual foundation of work they then write up and take credit for. “One could argue that the provost in this case is merely the minion doing the busy work for the AI, which then authors the paper,” she wrote — while Schnell argues the reverse, that the AI is the minion. Leidner is not buying it: “When one crafts thoughts into words, one is also shaping thoughts.”

Dartmouth says the technology must serve the educational mission, not the other way around. The open question — being litigated, awkwardly, in the pages of the student newspaper — is which way around it currently is. The detection business, at least, is thriving either way.