A transit auditor asks how the team proves AI-generated lab samples stayed controlled. What evidence matters most?
Select an answer to reveal the explanation.
Short Explanation
Auditors want a paper trail: stayed on the range, then destroyed—not 'we probably deleted it' vibes. Chat theme screenshots and vendor marketing do not count as custody.
Full Explanation
Transit auditors asking how the team proves AI-generated lab samples stayed controlled need custody evidence: logs showing samples remained inside the range and were destroyed after the test per policy. Chain-of-custody records (who generated, where stored, who accessed, when destroyed, and the destruction method) demonstrate control far better than informal claims. That evidence supports compliance, reduces dual-use leakage risk, and closes the audit loop on AI-assisted offensive artifacts used in municipal exercises.
Screenshots of the AI chat UI theme settings prove nothing about sample control. Verbal assurance that someone 'probably deleted' the files is weak and non-auditable. A marketing slide claiming the vendor is trusted does not document municipal custody of generated samples or their destruction timeline.
Exam caveat: what matters most is custody logs of range containment and post-test destruction—not UI aesthetics, verbal maybes, or vendor marketing claims.
Operational check: implement a sample register with timestamps, hash inventory, storage location (range-only), access list, and witnessed destruction; retain logs for the auditor's retention period; refuse to close the exercise ticket until destruction is recorded and verified.