[REVISION HISTORY]

Dartmouth College AI misconduct investigation

3 recorded revisions of this situation.

What changed

2026-11-01 01:16 UTC → 2026-11-03 02:36 UTC · added ~~removed~~

Dartmouth College launched an investigation into Provost Santiago Schnell after reports suggested several of his recent publications were written using artificial intelligence. The scrutiny followed an essay Schnell wrote for The Washington Post that was flagged by detection software. Student journalists reported that a review of nine of his works published that year showed a median AI-written score of 96%. Schnell disputed claims of misconduct, stating he used ChatGPT to refine arguments and check grammar, though he apologized for failing to disclose the use of AI. In response, President Sian Leah Beilock formed an independent advisory committee to review Schnell’s use of AI in academic and public writing. A faculty-led investigation, which reviewed university AI logs and emails, concluded that Schnell’s actions violated publication disclosure requirements and fell short of the institution’s ethics and professionalism standards. The review determined that Schnell used AI to ‘meaningfully alter his writing’ and failed to disclose this in several publications, including academic papers and a Washington Post opinion piece. Following these findings, Schnell resigned from his position. President Beilock stated the resignation was Schnell’s decision, made out of respect for his colleagues and the institution. Matthew Delmont has been appointed to serve as interim provost through June 2028. ~~The incident has sparked debate among academics regarding the regulation of AI in higher education, with some scholars questioning~~ Details from the ~~severity of~~ advisory committee revealed that Schnell provided insufficient cooperation during the probe. While he surrendered his university ~~response given~~ laptop, he declined to meet with investigators and did not provide logs from personal AI accounts. However, investigators accessed his Claude enterprise account, which showed he used the ~~ongoing uncertainty surrounding institutional norms for~~ chatbot to revise papers, fix code, and generate figures. In one instance, Schnell used the AI ~~integration.~~ to act as a specialist peer reviewer for a paper intended for the journal ChemSystemsChem.

Versions

Select a version to see what changed compared to the one before it.