AI Agent Security Solutions

Lasso Security AI agent security

Lasso Security says it discovers agents across Vertex AI, Microsoft Copilot, AWS Bedrock and Salesforce Agentforce through read-only integrations, and governs endpoint coding agents and MCP connections agentlessly, with Claude Code lifecycle hooks via enterprise management. Gateway mode blocks or masks traffic and integrates with Kong, Envoy AI Gateway, LiteLLM and other gateways. The vendor says it blocks or quarantines agent actions and out-of-scope tool calls before execution, streams audit trails to a SIEM and has an EDR integration. A third party reports $37 million raised. Browser-agent control, agent credential revocation, a price and public documentation: none found on 16 Nov 2026.

Facts read on .

Strength and weakness for AI Agent Security Solutions

Strength: Covers cloud agent builders, endpoint coding agents, gateways and audit with named integrations (lasso-security-32, lasso-security-16, lasso-security-26).

Weakness: No public docs or price, and no browser-agent or agent-credential revocation controls found (lasso-security-52, lasso-security-51, lasso-security-35; none found on 2026-11-16).

Scores and evidence

Lasso Security is joint 8th in the overall ranking at 58.4 of 100. Each score below lists the fact ids behind it; the facts follow further down with their sources and read dates.

Lasso Security: score per criterion
CriterionScoreWeightWhyFacts
Coverage: endpoint layer6 of 106The vendor inventories named coding agents and MCP connections agentlessly; local enforcement detail and OS coverage are limited.lasso-security-13, lasso-security-19, lasso-security-26, lasso-security-27
Coverage: browser layer2 of 106A usage-control extension exists; browser agent or AI browser coverage none found.lasso-security-36, lasso-security-37
Coverage: cloud workspace layer6 of 106Discovery across Vertex, Copilot, Bedrock and Agentforce via read-only integrations is described; runtime control in those platforms is limited.lasso-security-32, lasso-security-33, lasso-security-45
Coverage: identity layer3 of 106Role-based MCP access and over-privileged agent flags exist; credential inventory and revocation none found.lasso-security-24, lasso-security-34, lasso-security-35
Coverage: gateway layer7 of 106Gateway mode blocks or masks traffic and integrates with named AI gateways and a firewall; vendor claims only.lasso-security-14, lasso-security-16, lasso-security-44
Stops the action7 of 1020The vendor blocks or quarantines agent actions and out-of-scope coding-agent tool calls before execution; vendor claims only.lasso-security-10, lasso-security-12, lasso-security-14
Visibility7 of 1015Agent discovery, execution graphs, connection maps and tool-call audit trails are described; vendor claims only.lasso-security-25, lasso-security-32, lasso-security-38, lasso-security-39
Deployment7 of 1010Agentless, API, gateway and extension modes starting in hours are described; vendor claims only.lasso-security-33, lasso-security-36, lasso-security-46, lasso-security-47, lasso-security-48
Policy and governance6 of 1010Audit trails, framework mapping, SIEM streaming and an EDR integration are described; ticketing and IdP integrations none found.lasso-security-11, lasso-security-25, lasso-security-40, lasso-security-43, lasso-security-42
Transparency3 of 1010Research is published; a price and public documentation none found.lasso-security-49, lasso-security-51, lasso-security-52
Proof7 of 105A third party reports $37 million raised and names BMW, eToro and DHS as customers; analyst coverage is vendor-cited.lasso-security-02, lasso-security-03, lasso-security-04

Where it ranks

Product, deployment, price and owner

Facts and sources