CrowdStrike Secures AI Agents Inside Claude Cowork

When Claude Cowork executes code locally, it runs inside an isolated virtual machine on the user's computer. On the host, that VM runs as a single opaque process: The agent's activity inside the VM, including the processes it spawns, the packages it installs, and the network connections it makes, is not visible to the host's endpoint security sensor. Because host folders can be mounted into the VM, sensitive data can sit within the agent's reach.

Today at Fal.Con Europe 2026, CrowdStrike announced it is bringing CrowdStrike Falcon® Guardian protection natively into Anthropic's Claude Cowork. Built in partnership with Anthropic in response to enterprise customer demand, this integration extends Falcon protection to AI agents across their lifecycle, from agent creation through job execution. It runs the Falcon sensor directly inside the Cowork VM on both macOS and Windows to bring runtime visibility and detection into the agent's execution environment. 

The integration is now available in beta on macOS, with Windows support expected later in November.

How It Works

Consider this scenario: A user asks Cowork for a script to align their slides to a brand standard. While completing the task, the agent installs the latest version of a common Python package. Neither the user nor the agent knows that an adversary has compromised the package at its source. Once installed, that package reads host data through a mounted folder and sends it out to a known malicious endpoint. The Falcon sensor running inside the VM would detect it. This is exactly the type of malicious activity and potential data exfiltration the integration is built to detect as agents execute tasks. 

CrowdStrike’s sensor integration with the Cowork VM will deliver:

• Visibility and detection where agents execute: Process, package, and network activity inside the Cowork VM will no longer be a blind spot, so teams can detect and investigate threats with the same Falcon telemetry they use everywhere else.

• The same visibility on macOS and Windows: The VM runs the same Linux environment on both, so customers get consistent coverage regardless of the OS.

• Deployment and control through the CrowdStrike Falcon® platform: Admins will turn the integration on through their existing device management, the same way they deploy the Falcon sensor today, and the Falcon Linux sensor deploys inside the Cowork VM. From there, the team controls it through the Falcon platform so Claude Cowork comes under the same security control plane as the rest of the environment.

Built with Anthropic

CrowdStrike and Anthropic worked together to bring Falcon protection directly into Claude Cowork's execution environment. Anthropic enabled Falcon deployment within Cowork, while CrowdStrike engineered and tested the integration across installation, real-time protection, and ongoing updates. Through this beta integration, enterprises can extend the security they already trust from the Falcon platform to AI agents in Claude Cowork, alongside the virtual machine, permissions, and admin controls built into Cowork.

CrowdStrike’s Architectural Difference

AI agents don't operate in one layer, and neither does Falcon Guardian. Agents reach out over the network to communicate with LLMs, but they also execute processes locally on the workstations where they’re installed. These actions may never cross network boundaries. The Falcon sensor and Falcon Guardian see both agentic workstreams and fuse them together in a single view from CrowdStrike’s endpoint network inspection and OS-level visibility. This unified view, CrowdStrike® Agent Graph, shows analysts not only what agents were asked to do, but what they did at runtime. 

CrowdStrike also delivers agent-level enforcement controls via hook integrations for agents such as Claude Code, OpenAI Codex, GitHub Copilot, OpenCode, and Cursor.

The Claude Cowork integration extends Falcon Guardian coverage to a new place agents work. It takes Falcon Guardian’s runtime security foundation into the VM where Cowork runs its code, providing visibility and detection for what the agent executes, installs, and connects to across its lifecycle, from agent creation through job execution.

What's Next

Endpoint agents like Claude Cowork run their code inside a VM that host-only security can't see into. By running the Falcon Linux sensor inside the Cowork VM on macOS and Windows, this integration will close that gap. It will bring Falcon visibility and detection to where agents execute, it will be deployed and controlled through Falcon, and it will extend the runtime foundation of the three layers of AI security in Falcon Guardian.

The integration is focused on visibility and detection inside the VM. Available in beta on macOS, with Windows support expected later in November 2026, it enables customers to deploy the Falcon Linux sensor inside the Cowork VM through their existing device management tools. Their teams gain visibility and detection across the processes, packages, and network activity inside the VM.

Forward-Looking Statements

This release includes discussion of unreleased services or features. Any unreleased services or features referenced here are still in development and subject to change. Customers should make their purchase decisions based upon features that are currently available.