# Preprint finds single-run agent audits can miss irreversible damage

_Tuesday, August 18, 2026 at 12:00 AM EDT · AI, Science · Latest · Tier 2 — Notable_

A preprint introducing AgentRelBench reports that irreversible database-state damage occurred across the measured model families but not on every run, making one-shot audits unreliable for detecting some harmful agent behavior. Across 2,128 runs, the authors found a clean run missed a damage-producing model-task pair 0.80 of the time in the development pool. The held-out result was described as underpowered, and the capability gradient was observational rather than causal.

## Sources

- [cs.AI updates on arXiv.org](https://arxiv.org/abs/2608.15286)

---
Canonical: https://techandbusiness.org/newswire/HTInZQa81OKzymZILNrOIb
Retrieved: 2026-08-18T11:41:40.256Z
Publisher: Tech & Business (techandbusiness.org)