How We Evaluate Model Risk

A framework for measuring the security risk of LLMs by evaluating them as agents against adaptive attackers. Published on the Snyk Labs research blog.

1 min read   ·   August 19, 2026

2026   ·   ai-security   llm-agents   evaluation   ·   research