FrontierBreach
A model-driven security benchmark for measuring attack susceptibility in enterprise AI agents.
Is this your creation?

Description
A security benchmark that evaluates frontier model susceptibility to indirect prompt injection within complex, long-horizon enterprise workflows. It utilizes a four-stage, model-driven synthetic data generation pipeline to translate real-world threat intelligence into adversarial scenarios involving data exfiltration, ransomware, and arbitrary code execution.
Descriptions, tags, and model credits may be AI-generated or inferred from public sources and can be incomplete or wrong. Learn more.