Tel Aviv-based Bloom Security, which develops endpoint security tools for monitoring AI agents, extensions, and more, emerges from stealth with a $20M seed
Context & Ripple Effects
Bloom Security’s launch puts endpoint monitoring of AI agents and extensions into the security-tooling conversation. It follows nearby coverage of Bold’s AI-agent approach to securing enterprise devices, indicating that AI is becoming both a security capability and a new object of endpoint defense.
The company also arrives alongside Act Security’s push to reduce cloud access exposure, while Native has focused on monitoring security across cloud providers. Together, these developments point to security teams confronting access and visibility gaps across endpoints, cloud environments, and AI-enabled tools.
First-order effects
- Bloom Security gains seed funding to build and sell endpoint-security tooling aimed at monitoring AI agents, extensions, and related activity.
- Enterprise security teams evaluating AI-agent use gain another specialist option for visibility at the endpoint layer, rather than treating agents solely as a cloud or application-security issue.
Second-order effects
- Endpoint, cloud-access, and runtime-security vendors will face pressure to show how their controls cover AI-agent activity and extension risk across overlapping environments.
- Buyers may need to reconcile new AI-specific monitoring tools with existing endpoint and cloud-security workflows, raising the importance of integrations and clear ownership of alerts.
Third-order effects
- If AI agents and extensions become routine enterprise tools, the trusted-tool boundary will extend beyond traditional managed software to include agent permissions, actions, and connected extensions.
- The market could increasingly organize around an AI enforcement surface spanning devices, cloud access, and agent behavior, though it remains unclear whether specialists or broader security platforms will own that control point.
The trend: Enterprise security is expanding from protecting infrastructure and endpoints to governing the permissions and behavior of AI agents and the tools connected to them.