HazardAuditor: From Executable Threats to Safer Computer-Use Agents
Quick summary
arXiv:2609.15134v1 Announce Type: new Abstract: Computer-use agents increasingly interact with browsers, terminals, file systems, and external services, introducing safety risks that emerge through runtime behavior rather than generated content alone. Existing guard models target static prompts and responses and are poorly suited to agent execution; existing executable safety platforms produce evaluation verdicts rather than the normalized supervision a guard model needs to learn across heterogeneous agent frameworks. We introduce HazardAuditor, an execution-grounded framework that closes both
Key takeaways
- arXiv:2609.15134v1 Announce Type: new Abstract: Computer-use agents increasingly interact with browsers, terminals, file systems, and external services, introducing safety risks that emerge through runtime behavior rather than generated content alone.
- Existing guard models target static prompts and responses and are poorly suited to agent execution; existing executable safety platforms produce evaluation verdicts rather than the normalized supervision a guard model needs to learn across heterogeneous agent frameworks.
- We introduce HazardAuditor, an execution-grounded framework that closes both
Why it matters
“HazardAuditor: From Executable Threats to Safer Computer-Use Agents” shows why AI risk cannot be reduced to answer accuracy. Access controls, logging, human approval and incident response need to be designed into the workflow from the start.
Member comments