Washington – November 03, 2026 -- Evoke Security will integrate its agent detection and response platform with Claude Cowork, extending monitoring to AI agents that keep working in the cloud after employees close their laptops. The company announced the plan as Anthropic expands security integrations for Claude Enterprise. The integration is scheduled for a future release, and no date was given.

Agents move work off the endpoint, where security teams built a decade of visibility

Employees now hand AI agents real tasks: reading files, calling tools and handling sensitive data on their behalf. As that work shifts to the cloud, it leaves the endpoint, where security teams have spent a decade building visibility.

An agent can complete a legitimate task and still take an unexpected action along the way. Without a full record of what the agent did, security teams cannot tell routine agent work from the opening move of an incident.

Evoke Co-Founder and CEO Jason Rebholz said the first question in incident response is always "what happened?" He argued that an agent working somewhere security teams cannot see leaves that question unanswered, and that deeper Claude Cowork coverage inside the virtual environment closes the gap.

Evoke tracks the full agent trajectory and can block unauthorized actions

Evoke says existing security tooling was not built for how agents operate. Without agent-specific detection and response, teams may struggle to separate normal agent operations from risky behavior.

The platform monitors the full agent trajectory with a detection-in-depth approach. It identifies risky behavior, then either steers the agent to a safer path or blocks unauthorized actions.

The Claude Cowork integration builds on Evoke's broader scope: discovering agents running across an organization, assessing their exposure, and detecting and responding to the agent activity that matters. The company points to its recent Rogue Agent Assessment as evidence of how these risks surface in real environments today.