Noma Brings Agent Security to Employee Devices

Noma Security is extending agent governance to a place where many companies may have blind spots: employee laptops.

What happened

The company expanded its platform to discover and govern AI agents, MCP servers and agent skills running on employee devices.

Security teams can inspect permissions and credentials, set approved or blocked status and detect behaviours such as prompt injection, exposed secrets and excessive access.

The product is aimed at the growing use of local coding agents and productivity agents inside companies.

Why it matters

Agents running on employee machines can inherit the permissions of the people using them.

That creates a security problem outside centrally managed cloud applications. A local agent may access files, credentials or internal systems in ways that security teams cannot easily see.

Endpoint-level visibility is therefore becoming part of agent governance.

The bigger picture

AI adoption is moving faster than enterprise controls.

As employees install agents and connect them to tools, companies need ways to discover what is running, what it can access and whether it is behaving safely. Agent security is becoming an endpoint problem as well as a cloud problem.