July 1, 2026, (Inside AI) — Amazon Web Services has made WorkSpaces for agents generally available, giving AI agents a secure, managed cloud desktop to operate legacy business applications just like human users. The service targets enterprises stuck with critical desktop software—ERP systems, CRMs, mainframes, and proprietary tools—that are too customized, undocumented, or compliance-bound to modernize or abandon. Instead of building custom integrations, AI agents now see the screen and interact with these applications through a familiar WorkSpaces environment.
The launch addresses a persistent enterprise pain point: decades-old desktop applications that run core operations but resist API-based automation. WorkSpaces for agents uses the same infrastructure that organizations have relied on for over a decade to deliver secure, managed desktops at scale. Agents inherit identical identity controls, network isolation, and compliance boundaries as human users, so automation does not weaken governance. Early use cases include claims processing, patient record updates, trade settlement, and back-office operations.
The service works with any agent framework that supports the Model Context Protocol (MCP), an open standard for tool integration. Pricing scales on active session time, aligning cost with actual usage. Since its preview launch, customer and partner feedback has driven several new capabilities now available in general availability.
MCP tool forwarding lets agents interact with applications and the desktop operating system through direct MCP calls, bypassing traditional computer-use tools. This approach improves accuracy, reduces latency, and lowers cost. Real-time session control gives operators live visibility into agent activity and the power to revoke access mid-session, adding a safety layer for sensitive workflows. Domain-joined fleet support allows agents to operate under existing Active Directory identities, extending the same access policies and audit attribution that apply to employees.
Read: MCP 2026-07-28 Spec Moves to Stateless Core, Rolls Out Across Claude
These features reflect a careful balance between automation speed and enterprise control. By keeping agents inside familiar security perimeters, AWS aims to reassure compliance officers while giving developers a straightforward path to automate legacy processes. The direct MCP integration also signals a shift away from brittle screen-scraping or fragile UI automation toward more reliable, protocol-driven interaction.
Read: AWS Launches InvokeGuardrailChecks API for Agentic AI Safety
Developers can find documentation and sample code on GitHub, and the service is available now through the AWS console. As enterprises face mounting pressure to do more with less, WorkSpaces for agents offers a pragmatic bridge between old software and new AI capabilities—without ripping out the systems that still run the business.
