Lasso Details AI Security Platform for Agent Discovery, Testing and Runtime Guardrails

Summary

Lasso describes a platform that inventories AI agents, assesses configurations, red teams agent behavior, and applies runtime guardrails. The vendor says its Intent Security Engine detects intent misalignment and behavioral anomalies.

Key points

• Lasso says its AI Bill of Materials inventories agents across cloud environments, code repositories, and low-code or no-code platforms.
• The company says its AI-SPM maps agent relationships and identifies configuration risks against OWASP, NIST, and MITRE frameworks.
• Lasso describes static, multi-turn dynamic, and autonomous high-agency red-team testing, with findings feeding into runtime guardrail updates.
• The platform’s Intent Security Engine is described as detecting misalignment within an execution chain and deviations from an agent’s behavioral baseline.
• Lasso says inline guardrails can block risky actions in under 50 milliseconds; this is a vendor claim, not an independently verified result.
• The article reports vendor-stated detection accuracy of 98.6% and a 1.4% false-positive rate, without providing testing methodology.

People

Vendors

Products

Industries