ST-WebAgentBench

Evals

Benchmark / dataset

Safety and trustworthiness of enterprise web agents (policy compliance).

Category
Agent safety & harmful actions
Insurance exposure
Agent actions
Publisher
IBM Research
Publisher type
Big tech
License
Apache-2.0
Status
Maintained (<12mo)
Last commit
2026-03-12
Stars
29
Notes
Closest to enterprise policy breach.

View source on GitHub →

Wanich Risk Lab, 2026.