Dartmouth’s AI Champion Fell to the Disclosure Rules He Urged Others to Navigate
Dartmouth’s AI Champion Fell to the Disclosure Rules He Urged Others to Navigate
Santiago Schnell arrived at Dartmouth’s AI controversy as one of its most visible participants. The provost, who had held the university’s top academic post since July 2025, oversaw work at an institution that proudly traces the birth of artificial intelligence to its 1956 summer workshop.
In August 2026, Schnell published a Washington Post op-ed titled, “Universities are fighting AI cheating. But there’s a deeper problem.” He later told Semafor he had used ChatGPT to “refine a few arguments, check for grammatical errors,” and prepare the piece for submission. His broader view was that AI could assist scholarship without becoming its author: in a later paper, he argued that “contribution and authorship are different roles.”
But Dartmouth’s faculty advisory committee found a wider and insufficiently disclosed role for the technology. Reviewing logs from the university’s Claude account, it said Schnell used the tool to substantially edit a ChemSystemsChem manuscript, troubleshoot code and generate figures—activity the journal required authors to disclose. The committee also concluded that his Post disclosure fell short of the newspaper’s standards, and that he delayed or failed to answer requests during the review.
The gap between Schnell’s account of editorial assistance and the committee’s assessment became central. He did surrender his Dartmouth laptop, but declined to provide logs from personal LLM accounts and would not meet with investigators; the report described his cooperation as insufficient and said his public explanations could understate his actual AI use.
This week, Schnell resigned. President Sian Leah Beilock said the university would continue pushing to remain at the forefront of AI, while stressing that leaders must “model the transparency and human judgment that we expect from our students and scholars.” The episode leaves Dartmouth with a stark lesson: experimental technology may be fluid, but disclosure rules are not optional.
Write a comment