Dartmouth Provost Resigns Over Undisclosed AI Usage
Academic integrity failure leads to resignation after undisclosed LLM use in published works.
1Provost Resigns Following Investigation
Santiago Schnell, the provost of Dartmouth College, resigned this week after an advisory committee investigation. The probe focused on his failure to disclose the use of generative AI in an academic paper and a Washington Post op-ed. These publications explicitly required authors to declare any AI assistance during the drafting process.
2Contradictory Public Stance
On August 9, 2026, Schnell authored an op-ed in the Washington Post titled 'Universities are fighting AI cheating. But there’s a deeper problem.' Ironically, his own resignation stems from the exact issue he publicly addressed. His failure to adhere to disclosure policies undermined his credibility regarding institutional academic integrity and AI governance.
3Insufficient Cooperation During Probe
The advisory committee report explicitly cited 'insufficient cooperation' from Schnell during the investigation. While he surrendered his university-issued laptop, he refused to provide logs from his personal LLM accounts. Furthermore, he declined to meet with the committee, which significantly hindered their ability to fully audit the extent of his AI usage.
4Evidence from Enterprise Accounts
Investigators successfully accessed Schnell’s Claude enterprise account provided by Dartmouth. The logs revealed that he utilized the LLM for various tasks, including revising academic papers, identifying and fixing bugs in code, and generating figures. This digital footprint provided the committee with concrete evidence of his reliance on AI tools for professional work.
5Institutional Disclosure Requirements
Academic publishers and institutions are increasingly mandating transparency regarding AI-assisted content creation. By failing to disclose his use of Claude, Schnell violated these established ethical standards. The committee’s findings highlight that even high-ranking administrators are subject to the same rigorous disclosure protocols expected of faculty and students within the academic community.
6The Mechanics of AI Assistance
The investigation detailed how Schnell used Claude to refine his writing and technical output. By asking the model to process his papers, he bypassed the requirement for original authorship transparency. This case demonstrates how easily LLMs can be integrated into academic workflows, creating significant challenges for institutions attempting to enforce strict integrity.
7Consequences of Opaque Practices
The resignation underscores the severe professional consequences of failing to maintain transparency in an era of generative AI. By withholding personal logs and refusing to cooperate, Schnell transformed a policy violation into a leadership crisis. His departure serves as a cautionary tale for administrators balancing technological adoption with institutional ethical obligations.
8Broader Implications for Academia
This incident highlights the tension between leveraging AI for productivity and maintaining academic rigor. As universities grapple with AI integration, leaders are expected to model the very transparency they demand from others. The case reinforces that institutional trust is fragile and hinges on the consistent application of disclosure rules across all levels.
Original article: Ars Technica - All content — Dartmouth Provost Resigns Over Undisclosed AI Usage