AgentHazard: A Benchmark for Evaluating Harmful Behavior in Computer-Use Agents
Computer-use agents extend language models from text generation to persistent action over tools, files, and execution environments. Unlike chat systems, they maintain state across interactions and translate intermediate outputs into concrete actions. This creates a distinct safety challenge in that harmful behavior may emerge through sequences of individually plausible steps, including intermediate actions that appear locally acceptable but collectively lead to unauthorized actions. We present \textbf{AgentHazard}, a benchmark for evaluating harmful behavior in computer-use agents. AgentHazard contains \textbf{2,653} instances spanning diverse risk categories and attack strategies. Each i...
| Authors | Yunhao Feng, Yifan Ding, Yingshui Tan et al. |
|---|---|
| Primary category | cs.AI |
| Published | 3 April 2026 |
| Last updated at source | 22 September 2026 |
Connected in the knowledge graph
- Yunhao-Feng/AgentHazardDataset · cited by
Models and datasets referencing this paper
- Yunhao-Feng/AgentHazardDataset
Silytics shows public metadata from arXiv and links to the original. Information may be incomplete or out of date. Last refreshed 17 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.