On August 9, 2026, Dartmouth College provost Santiago Schnell published a Washington Post op-ed warning universities about the dangers of artificial intelligence in student work. This week, Schnell resigned after an advisory committee determined he failed to disclose significant AI use in both an academic paper and that op-ed, despite publisher requirements for such transparency. The investigation noted insufficient cooperation from Schnell, who provided his university laptop but withheld logs from his personal large language model accounts. He also declined to meet with the committee during the inquiry.
The committee accessed Schnell’s enterprise Claude account used through Dartmouth, revealing he employed the tool to revise papers, debug code, and generate figures. Specific prompts show Schnell asking the model to fix errors and edit diagrams within his scholarly output. This case highlights the gap between public warnings on academic integrity and the undisclosed reliance on generative tools by senior administrators. The incident underscores the difficulty in enforcing disclosure policies when officials do not share their own digital records.
- Schnell resigned after failing to turn over personal LLM logs.
- The committee found undisclosed AI use in two major publications.
- Enterprise account logs confirmed extensive tool usage for writing support.
