Dartmouth provost resigns after AI use in published work comes under scrutiny

Dartmouth provost Santiago Schnell has resigned after an investigation found AI use in several published works, according to Fox News. The case puts a practical question about AI and academic integrity in a senior university office, rather than only in students’ classrooms.

Watch Desk analysis

What happened

Fox News says Dartmouth’s faculty-led committee reviewed internal emails and university records and identified two additional academic publications in which Schnell used AI to varying degrees. In one, it says, he used the university’s enterprise Claude tool to substantially edit a research manuscript, troubleshoot code and generate figures. The committee also found that his August opinion piece in The Washington Post violated that paper’s stated AI-use policy.

Fox News says President Sian Leah Beilock announced the resignation on Wednesday and described it as Schnell’s decision. The report also recounts Semafor’s earlier claim that an AI-detection tool judged the op-ed to be entirely AI-written. That detector-based claim is not the same as the committee’s reported findings, and should not be treated as proof of who wrote the piece.

Why it matters

The reported findings go beyond AI-assisted prose: they include code troubleshooting and figure generation in academic work. That makes the case relevant to how universities define acceptable assistance, disclose it and assess whether published work remains the author’s own.

The distinction between using a tool and misrepresenting its use matters here. Fox News says the committee found a breach of The Washington Post’s policy, but the supplied report does not establish that every use of AI in the academic publications violated university rules.

Our read

This is a consequential example of AI-use rules reaching beyond student assignments and into research and university leadership. The reported committee review gives the story more substance than a detector score alone. Still, a claim that text is “100% AI-written” deserves particular care: detection software is not authorship evidence by itself, and the committee findings should be kept separate from Semafor’s earlier assessment.

What to watch

• Whether Dartmouth publishes the committee’s findings or explains its rules for AI-assisted research.
• Whether The Washington Post or Schnell offers further detail on the op-ed and the policy breach.
• How universities distinguish permitted assistance from work that must be disclosed or cannot be delegated.

Discussion spark: Should universities treat AI use in published research as acceptable assistance when it is disclosed, or should some tasks, such as generating figures or troubleshooting code, require stricter limits?

Watch Desk is operated by WittyWires as an independent cross-cutting AI news tracker. It does not speak for the organisations or people it covers.