Dartmouth Provost Resignation Exposes AI Disclosure Policy Enforcement Gap
Source
Dartmouth's provost warned us about AI cheating - but his own AI use cost him his jobDartmouth College
What happened
Dartmouth College provost Santiago Schnell resigned on October 30, 2026. An advisory committee concluded he had used AI tools to assist in writing a published academic paper and a Washington Post op-ed. He did not disclose that assistance, as both publishers required. The committee's report, as covered by Ars Technica, noted that Schnell did not cooperate fully with the investigation. He declined to provide logs from personal AI accounts and did not agree to meet with investigators. Logs from his Dartmouth-managed Claude enterprise account were reviewed, but investigators could not access records from any personal accounts he may have used. Schnell had publicly warned about AI misuse in academic contexts. This made the nondisclosure a direct accountability incident rather than a policy ambiguity case.
Why it matters
- ·Enterprise AI use policies that require disclosure of AI assistance in written work are unenforceable when employees use personal AI accounts rather than employer-managed tools. This case shows that audit logs from enterprise accounts alone cannot close the gap.
- ·The investigation illustrates how the NIST AI Risk Management Framework (AI RMF 1.0) and Playbook principle of accountability depends on verifiable records. Organizations whose AI policies rely on self-attestation rather than log-based evidence face the same evidentiary problem Dartmouth encountered.
- ·For organizations in regulated industries or those subject to disclosure obligations, this incident is a signal. AI use in externally published work, such as regulatory submissions, research reports, or public communications, requires controls beyond policy statements. Log access and verification are necessary.
Stay ahead of stories like this
Get every US AI governance development like this one, plus the rest of the week's developments. Every Thursday.
Recent issues
- Three separate incidents this week each created a new compliance action item: verify your AI vendors' data sourcing, demand agent monitoring coverage guarantees, and audit trusted app identities in your cloud environment.29 Oct
- Regulators, litigants, and attackers are all moving faster than most organizations' AI governance frameworks, shifting leverage away from deployers and toward those who can prove control, accountability, and honest disclosure.21 Oct
Free every Thursday. Unsubscribe anytime.
Governance controls affected
What to do now
- ☐Map every context in which your AI use policy requires disclosure of AI assistance in written work, and identify whether those disclosures are currently self-attested or log-verified.
- ☐Ask your IT team which AI tools used by employees are covered by enterprise account logs and which are personal accounts outside your visibility. Identify the gap.
- ☐Review your AI use policy to determine whether it restricts or requires employees to use only employer-managed AI accounts for work that will be published, submitted to regulators, or used in official communications.
- ☐Confirm that enterprise AI account logs are retained long enough to support a policy investigation, and test whether those logs are accessible to HR, legal, or compliance teams when needed.
- ☐Assess whether your policy includes consequences for nondisclosure of AI assistance and whether those consequences are communicated clearly to staff at all levels, including leadership.
What to watch next
Publisher and journal disclosure requirements for AI use are becoming more specific, and enforcement through institutional accountability processes is now a documented outcome rather than a theoretical risk. Compliance teams should monitor whether any regulatory bodies in their sector, such as financial regulators or healthcare agencies, adopt parallel disclosure requirements for AI assistance in submissions and reports. The gap between enterprise-managed AI account logs and personal account use may attract regulator attention. This includes proposals under discussion in the United States and the European Union under the EU AI Act (Regulation (EU) 2024/1689).