Authorization-Grounded Workflow Attacks on Tool-Using Agents

# Authorization-Grounded Workflow Attacks on Tool-Using Agents

https://gcdradar.vercel.app/incidents/authorization-grounded-workflow-attacks-on-tool-using-agents · AIR-2026-1395 · closed 2026-11-02

- Lane: Research
- Severity: 1 (Research / proof of concept)
- Fix status: Fix status unknown
- First reported: 2026-10-19
- AI role: Attack on AI
- Products: OpenClaw

## Summary

Researchers demonstrated a workflow attack against tool-using language model agents that bypasses explicit policy overrides by routing through coherent approval steps. The technique was evaluated as part of the OpenClaw Agent Security Attack Competition.

## What to do

No action recorded; read the sources below and the vendor's advisory.

## Latest updates

- 2026-10-19: When Approval Becomes Execution: Authorization-Grounded Workflow Attacks on Tool-Using Agents (https://arxiv.org/abs/2610.23077)

## Sources

- 2026-10-19 arXiv: When Approval Becomes Execution: Authorization-Grounded Workflow Attacks on Tool-Using Agents https://arxiv.org/abs/2610.23077

Verify against the vendor's advisory before changing production. Data CC BY-SA 4.0, https://gcdradar.vercel.app.