Enterprise AI security company

Verifiable control for AI security research

An emerging leader in enterprise AI security required deep, queryable visibility to match its Zero Trust philosophy.

At a glance

• Endpoints: ~35 (macOS and Linux).
• Primary requirement: security-as-code and OS interoperability.
• Key integrations: CI/CD pipelines.
• Previous solution: legacy MDMs.

The challenge: "black box" visibility

They found legacy MDMs to be restrictive "black boxes" that lacked necessary Linux visibility.

The solution: Linux visibility at scale

Fleet’s use of osquery turned Linux workstations from a blind spot into a source of real-time security telemetry. The open-source core allows them to audit their own management infrastructure, which is a key requirement for their security-first mission.

The results: instant CVE response

• Zero-impact transition: engineers appreciated the move to a lightweight agent that doesn't hinder productivity.
• IDE security: they use Fleet to monitor for risky IDE extensions (like VS Code or Cursor) to ensure AI coding tools remain secure.
• SOC 2 maintenance: when a new CVE is announced, they run a live query to get an answer in seconds, which is vital for maintaining SOC 2 status.