Praetorian Launches Augustus To Test LLM Security

Praetorian on Feb. 10, 2026 released Augustus, an open-source vulnerability scanner designed to secure large language models (LLMs) against adversarial threats. Augustus ships as a single-binary tool that can launch more than 210 distinct adversarial attacks against 28 LLM providers, bridging academic research tooling and production-grade security testing. Security teams can use Augustus to run comprehensive adversarial evaluations locally or integrate it into CI pipelines.