What happened
The Dartmouth’s investigation revealed that several of Provost Santiago Schnell’s recent publications were detected as almost entirely AI‑written by the Pangram detector. A faculty‑led petition calling for his removal has gathered over 220 signatures. President Sian Leah Beilock announced an independent committee, led by former provost David Kotz, to review Schnell’s AI use. Schnell issued an apology for not consistently disclosing AI assistance.
In September 2026, The Dartmouth reported that a University of Chicago‑audited AI detector, Pangram, flagged multiple recent works by Provost Santiago Schnell as 92% to 99% AI‑written. The investigation, building on an earlier Semafor report that identified Schnell’s August 9 Washington Post op‑ed as 100% AI‑generated, sparked immediate concern among faculty, students, and alumni.
The Dartmouth chapter of the American Association of University Professors circulated a petition demanding Schnell’s removal, which quickly amassed more than 220 signatures from across the Dartmouth community. Schnell did not respond to a request for comment before publication.
College President Sian Leah Beilock announced the formation of an independent review committee, chaired by former provost and computer‑science professor David Kotz ’86, to conduct a “rigorous, objective review” of Schnell’s AI use in academic publications and public writing. Schnell later apologized in a letter to the editor, acknowledging he should have disclosed AI assistance more consistently.
Interviews with faculty and students revealed divergent views on what constitutes responsible AI use. Some, like biology department chair Magdalena Bezanilla, called for full disclosure of any AI involvement, while others, such as economics professor Paul Novosad, emphasized context‑dependent standards. The college’s existing guidelines grant professors discretion but require student acknowledgment of AI use, and interim dean John Carey recently sent faculty additional guidance on employing AI‑detection tools.
Source details: thedartmouth.com ↗
Why it matters
The case highlights the growing tension between capabilities and academic integrity standards. It raises questions about consistent disclosure expectations for faculty versus students, the adequacy of existing AI policies, and the potential erosion of trust if leaders are perceived to bypass rules. The outcome could shape how higher‑education institutions nationwide craft and enforce AI‑use guidelines.
The incident underscores the need for clear, enforceable AI disclosure policies in academia, where expectations for original authorship are central to scholarly credibility. Without consistent standards, faculty may gain unfair advantages, undermining student trust and the integrity of the honor code.
It also illustrates the practical challenges of relying on AI‑detection tools. While Pangram boasts a near‑zero error rate per a 2025 audit, its findings are still contested by some faculty, raising questions about the reliability of such technology for disciplinary actions.
The response from Dartmouth could set a precedent for other institutions grappling with similar issues, influencing how universities balance AI integration for research productivity with safeguards against undisclosed AI‑generated content.
Interactive Mechanism: How It Actually Works
Explore the underlying technology behind this development interactively.
Impossibility results in algorithmic fairness (e.g. Kleinberg et al., Chouldechova) show what?
What to watch next
The independent committee’s findings and any resulting policy revisions at Dartmouth; potential actions by the Board of Trustees regarding Schnell’s position; broader adoption of AI‑detection tools like Pangram across other universities; and how the incident influences national discussions on AI disclosure standards in scholarly publishing.
The final report from the independent committee and any subsequent disciplinary or policy actions against Schnell.
Potential revisions to Dartmouth’s guidelines, including whether mandatory disclosure becomes a formal requirement for all faculty publications.
Reactions from the Board of Trustees and whether the petition leads to formal proceedings to remove the provost.
Adoption of similar AI‑detection and disclosure frameworks by peer institutions, especially as tools become more prevalent in scholarly work.