Keeper Security has expanded its endpoint privilege management platform to include governance for AI agents, signalling a shift in how enterprises secure rapidly proliferating non-human identities across employee devices.
The new capability integrates agentic AI governance into Keeper Endpoint Privilege Manager, enabling organisations to discover, monitor and control AI agents operating on workstations in real time. The move positions the platform as a unified governance layer for both human users and autonomous software agents.
Closing the AI governance gap
The launch comes amid accelerating adoption of AI agents and widening governance gaps. Gartner forecasts that by 2028, a typical Fortune 500 enterprise will deploy more than 150,000 AI agents, up from fewer than 15 in 2025.
However, governance frameworks remain underdeveloped: IBM data indicates that 63% of organisations lack formal AI governance policies, while 97% of those experiencing AI-related breaches did not have adequate access controls in place.
Keeper argues that existing approaches fall short because they focus primarily on the Model Context Protocol (MCP) layer, governing only a subset of agent activity. Its platform instead operates at the operating system level, where agents execute actions, providing visibility into behaviours such as file access, privilege escalation and process creation.
By applying consistent policy enforcement across both human and machine identities, organisations gain a unified audit trail and clearer attribution of actions across endpoints.
From detection to enforcement
A core feature of the update is the ability to identify both known and unknown AI agents running on managed devices. Recognised tools such as GitHub Copilot, Claude Code and Amazon Q are matched against a verified catalogue, while unknown applications are assessed using a proprietary scoring system that flags potential AI agents based on behaviour.
Once identified, agents are automatically subjected to policy controls without requiring manual classification. Administrators can define thresholds and groupings, enabling tailored governance across different user roles and risk profiles.
Three policy layers underpin the framework: controls over who can run AI agents, what actions they can perform, and how they request elevated privileges. A “monitor-first” approach allows organisations to observe behaviour before enforcing restrictions, while sensitive actions can be routed for human approval.
“AI agents are not assistants; they are principals,” said Darren Guccione, CEO and Co-founder of Keeper Security. “If you are not governing them with the same rigour you apply to your human workforce, you have blind spots that adversaries will find before you do.”
Compliance and visibility
Craig Lurey, CTO and Co-founder, added that the technology addresses a rapidly emerging threat landscape. “AI agents operate with an alarming level of autonomy, creating an urgent security gap that organisations are scrambling to close.”
The platform also introduces enhanced dashboards and audit capabilities, giving security teams visibility into agent activity and policy decisions. The unified audit trail supports compliance with frameworks such as the NIST AI Risk Management Framework, enabling organisations to align AI adoption with regulatory expectations.
