P110 Personnel-based verification and whistleblowers
What certainty can personnel-based verification and whistleblower channels give, and can they survive a prover that compartmentalises, surveils or intimidates staff?
Personnel-based verification and whistleblower channels.
Status
Analyzed Whistleblower clause only: a draft treaty article (Article X) with anti-retaliation duties, confidential or anonymous reporting channels run by the treaty body, identity protection, and asylum for whistleblowers and their families; a proposal, untested. The certainty personnel-based verification can give: no work beyond Harack's stated limit ('achievable certainty might be much lower than ... with hardware'). Survival against staff under video surveillance or with backdoored computers: Baker et al. state the limit, no design. Reach into state-run operations abroad: Rahman states the limit.
Parties prohibit and prevent retaliation against Covered Whistleblowers and Associated Persons, including but not limited to dismissal, demotion, blacklisting, loss of benefits, harassment, intimidation, threats, civil or criminal actions, visa cancellation, physical violence, imprisonment, restriction of movement, or other adverse measures.
Posed by
- Harack et al. (2025), Verification for International AI Governance (Oxford AIGI):
there are significant challenges with robustly verifying these due to the asymmetric burden of proof for verifying negative claims
(§2.1.1, p.39, L2141–2143) (source states the problem but not the quantity the question asks for)
Also raised in: Baker, Kulp, Marks, Brundage & Heim (2025), Verifying International Agreements on AI: Six Layers of Verification for Rules on Large-Scale AI Development and Deployment (RAND WR-A4077-1); Scher & Thiergart (2024), Mechanisms to Verify International Agreements About AI Development (MIRI); Rahman (2026), Does Distributed Training Undermine Compute Governance? (MIRI); Wasil, Reed, Miller & Barnett (2024), Verification Methods for International AI Agreements.
Sub-questions
| ID | Sub-question | What the sources report |
|---|---|---|
| P110.1 | Can whistleblowing survive an adversary that physically/digitally blocks contact or intimidates employees, and can absence of intimidation be verified? | a source suggests an answer (not checked) |
| P110.2 | Can whistleblower programmes reach insiders at state-actor operations abroad? | open |
Sub-questions not yet stated precisely enough to check
| ID | Sub-question |
|---|---|
| P110.3 | What certainty can personnel-based verification give, given negative claims (that the monitored personnel are the only relevant ones) are hard to prove? |
| P110.4 | Does access to personnel (interviews) fail when a prover compartmentalizes and employs only loyal staff? |
| P110.5 | Can whistleblower channels protect sources who are under video surveillance or whose computers are backdoored? |
| P110.6 | What design of whistleblower programme works for AI compute verification? |
| P110.7 | How fast does personnel-based verification lose power as AI automation reduces headcount? |
| P110.8 | Can relevant personnel be tracked within privacy and sovereignty limits? |
Possibly relevant work found by automated search
Found by AI agents searching for this problem's sub-questions, mostly by keyword and citation in the OpenAlex scholarly database and on the web; a second AI pass reviewed each one, and we spot-checked some by hand. They seemed reasonably related, as far as one can tell from outside each paper's field. Some come from neighbouring fields, such as arms control and financial audit, and bear on the problem by analogy. Not reviewed against the status above; a paper listed here does not change that status. 928 of 1052 quotes checked word for word against the paper's text.
For P110.1: Can whistleblowing survive an adversary that physically/digitally blocks contact or intimidates employees, and can absence of intimidation be verified?
For P110.2: Can whistleblower programmes reach insiders at state-actor operations abroad?
For P110.3: What certainty can personnel-based verification give, given negative claims (that the monitored personnel are the only relevant ones) are hard to prove?
For P110.4: Does access to personnel (interviews) fail when a prover compartmentalizes and employs only loyal staff?
* Full text not reviewed; more likely to contain errors.
Data
- problems.csv — one row per problem
- problem_status.csv — status, scope and evidence per problem
- problem_sources.csv — the posed-by quote and other documents per problem
- leaves.csv — all sub-questions, including those without a checkable form
- leaf_sources.csv — every source row behind each sub-question