Executive Summary
- Provost Resignation. Dartmouth College Provost Santiago Schnell stepped down after a faculty inquiry found undisclosed AI use in research and public writing.
- Enterprise Tool Logs. Investigators reviewed campus server logs showing Schnell used Anthropic’s Claude to edit text, debug code, and make charts for a journal article without disclosure.
- Washington Post Op-Ed. Schnell admitted using OpenAI’s ChatGPT to refine arguments in an op-ed about college AI cheating, breaching newspaper policy.
- Birthplace of AI Scrutiny. The resignation carries symbolic weight at Dartmouth, where scholars first coined the term artificial intelligence in 1956.
- Interim Leadership. History professor Matthew Delmont will serve as interim provost through June 2028 as Dartmouth reviews academic transparency codes.
Dartmouth College Provost Santiago Schnell has resigned. His exit follows a faculty review of his writing. The panel found that Schnell failed to disclose using AI tools in scientific papers and public articles.
The findings carry deep symbolic weight. Dartmouth is known as the historic birthplace of AI. Scholars coined the term on campus in 1956. Now, the Ivy League school faces national scrutiny over leadership ethics. Schnell is the most prominent college leader to step down over hidden AI use.
Table of Contents
Enterprise Logs and Journal Submission Rules
The review began after questions arose about Schnell’s papers. Faculty members checked logs from Dartmouth’s enterprise AI tool. The records showed that Schnell used Anthropic’s Claude on a paper for the journal ChemSystemsChem.
Schnell used Claude to edit text, fix code bugs, and make charts. In one prompt, he asked the model if it saw any scientific flaw before submission. The journal required writers to report AI help. Schnell did not report his AI use at submission or publication.
The committee found that Schnell was not forthcoming about his AI use. He delayed answering requests and fell short of school ethical codes.
The Washington Post Op-Ed Controversy
Schnell also ran into trouble with his public writing. In August, he wrote an op-ed for The Washington Post. The piece discussed college fights against student AI cheating.
Schnell later told news site Semafor that he used ChatGPT to hone arguments and fix grammar. But The Washington Post bans undisclosed AI text. The faculty panel found that Schnell broke the paper’s rules. This contrast sparked intense campus debate, as colleges penalize undergraduates for similar practices.
Comparison Factsheet: Investigation Findings vs. Policy Standards
The factsheet below outlines the writing projects reviewed by the faculty committee, the AI tools involved, and the specific disclosure violations identified.
| Case Work | AI Tool | How AI Was Used | Stated Rule | Panel Finding |
|---|---|---|---|---|
| ChemSystemsChem Paper | Anthropic Claude | Edited text, fixed code, drew charts | Must report AI help when submitting | Violation: Did not report AI help |
| Washington Post Op-Ed | OpenAI ChatGPT | Honed points, checked grammar | Bans hidden AI text | Violation: Broke paper rules |
| Faculty Inquiry | Not Applicable | Replies to faculty review panel | Dartmouth Code of Conduct | Violation: Delayed answers to panel |
| Provost Role | Enterprise AI Logs | Top academic and budget leader | College leadership standards | Resigned in October 2026 |
AI Ethics at the Birthplace of Artificial Intelligence
Dartmouth occupies a unique place in computing history. In the summer of 1956, mathematician John McCarthy organized the Dartmouth Summer Research Project on Artificial Intelligence. The gathering coined the term artificial intelligence and founded the discipline. That legacy makes ethical missteps on campus especially sensitive.
Dartmouth President Sian Leah Beilock accepted Schnell’s resignation. She stressed that leaders must model the honesty expected of scholars and students. History professor Matthew Delmont will serve as interim provost through June 2028.
Higher Education’s Dilemma on Academic Transparency
The case highlights a growing dilemma across universities. As campus leaders deploy enterprise AI services, as seen in Cornell providing free campus AI tools, policies remain uneven. Students face strict scrutiny for using generative tools in coursework, while faculty guidelines are often ambiguous.
Schnell’s resignation shows that enterprise audit logs can reveal undisclosed AI use. As universities navigate generative tools, institutional credibility will depend on holding senior leaders to the same standards enforced in the classroom.
Explore Engineering Utilities and AI Solutions
Test payload formats, token calculations, and text transformations using our suite of Free Online Developer Tools.
Auditing institutional AI policies, setting up enterprise model governance, or managing compliance workflows? Consult with TeckScaler’s enterprise IT specialists →
Frequently Asked Questions
Why did Dartmouth Provost Santiago Schnell resign?
A faculty inquiry found that Schnell failed to disclose AI use in a scientific paper and a Washington Post op-ed. The panel ruled his conduct fell short of school ethical standards.
How did investigators discover the undisclosed AI use?
Investigators reviewed server logs from Dartmouth’s enterprise AI platform. The records showed specific prompts asking Claude to edit drafts, debug code, and make charts.
What AI tools did Santiago Schnell use in his writing?
Schnell used Anthropic’s Claude for a research paper in ChemSystemsChem. He also used OpenAI’s ChatGPT to polish an op-ed published in The Washington Post.
Who will serve as Dartmouth’s provost following the resignation?
History professor Matthew Delmont will serve as interim provost through June 2028. He previously served as an associate dean of faculty at the college.
[…] As universities expand enterprise access to AI tools, institutional integrity policies are coming under unprecedented scrutiny. This was demonstrated when a top Ivy League administrator stepped down following an inquiry into Dartmouth provost resignation over undisclosed AI writing. […]