Amazon WorkSpaces for AI agents now supports Linux and Windows 11 cloud desktops in addition to the existing Windows Server option. This expansion lets each AI agent run on the operating system its workflows and applications require, so agents can work in the same environments your workforce already uses without re-platforming.
Linux desktops give agents full root access for developer workflows
Linux desktops provide AI agents with full root access to the operating system where many developers already build, test, and automate agentic workflows. Agents can take on developer tasks such as running builds, tests, and scripts directly in the same Linux environment your engineering teams use daily. This capability means you can deploy agents that compile code, execute test suites, and run automation scripts without translating those workflows to a different OS.
The agent operates inside a familiar Linux desktop with the same toolchains, package managers, and shell environments your developers rely on. Because the desktop runs in WorkSpaces, you retain the enterprise-grade security and governance controls your organization already enforces. Network isolation, IAM-based access, and existing compliance boundaries apply to these Linux desktops just as they do for Windows Server instances.
You can assign Linux desktops to agents that need to interact with container runtimes, language-specific build systems, or infrastructure-as-code tooling that expects a Linux host. The root access lets agents install packages, modify system configuration, and manage services exactly as a human developer would, while the WorkSpaces control plane ensures the underlying infrastructure remains governed and auditable.
Windows 11 desktops let agents work in everyday enterprise applications
Windows 11 desktops enable AI agents to operate inside the same enterprise applications your teams depend on every day. Agents can interact with line-of-business apps, productivity suites, and custom Windows software without requiring those applications to be re-platformed or web-enabled. This is particularly valuable for workflows that involve GUI-driven tools, legacy Windows applications, or software that only runs on modern Windows client SKUs.
The agent sees a standard Windows 11 desktop, complete with the same user profile, installed applications, and system configuration your employees use. As with Linux, these Windows 11 desktops inherit the same security and governance controls as the existing Windows Server offering. Network isolation, IAM-based access policies, and your organization's compliance boundaries remain consistent across all three supported operating systems.
You can deploy agents that automate data entry in ERP clients, generate reports from thick-client analytics tools, or drive workflows in custom .NET applications that have no API surface. The Windows 11 desktop provides the same registry, file system, and COM interfaces those applications expect, so the agent behaves like a local user session while running in a managed cloud desktop.
Unified security and governance across all three OS options
Linux, Windows 11, and Windows Server desktops for AI agents all share the same security and governance framework. This includes network isolation that keeps agent traffic segmented, IAM-based access control that ties agent permissions to roles, and the compliance boundaries your organization has already established for WorkSpaces. You can apply consistent policies regardless of which operating system an agent runs on.
For example, the same IAM condition keys, VPC configurations, and logging integrations work across Linux and Windows desktops, so your security team does not need to maintain separate control planes for each OS. This uniformity simplifies operations when you run a mixed fleet of agents — some on Linux for build and test automation, others on Windows 11 for business application interaction, and still others on Windows Server for legacy service compatibility.
The shared control plane also means audit logs, session recording, and data loss prevention policies apply uniformly. Whether an agent is compiling a Go binary on Ubuntu or clicking through a WPF forms application on Windows 11, the governance posture remains identical, reducing the operational overhead of managing heterogeneous agent workloads.
What to do next
To start running AI agents on Linux or Windows 11 WorkSpaces, review the agent access documentation and identify which of your agent workloads benefit from a Linux developer environment versus a Windows 11 enterprise desktop. Then provision the matching desktop type through the WorkSpaces console or API.
Source: Amazon WorkSpaces for AI agents now supports Linux and Windows 11 cloud desktops (AWS).
