Microsoft developing local AI agent to compete with open-source OpenClaw
Microsoft is testing OpenClaw-like features for Microsoft 365 Copilot aimed at enterprise customers, the company confirmed to The Information. The agent would run continuously to complete multi-step tasks over extended periods, distinguishing it from Microsoft's existing cloud-based agents like Copilot Cowork and Copilot Tasks.
Microsoft is developing an OpenClaw-like agent for Microsoft 365 Copilot that would target enterprise customers with enhanced security controls, the company confirmed to The Information.
The new agent aims to run continuously, executing multi-step tasks over extended periods. Microsoft has not confirmed whether the agent will run locally on user hardware like OpenClaw or adopt a hybrid approach.
How it differs from existing Microsoft agents
Microsoft has launched multiple AI agents in recent months:
Copilot Cowork (announced March 2026): Takes actions within Microsoft 365 apps, powered by Microsoft's "Work IQ" technology with Anthropic's Claude as an optional model. Runs in the cloud, not locally.
Copilot Tasks (preview launched February 2026): Designed for prosumer tasks including email organization and travel planning. Also cloud-based.
The new OpenClaw-style agent would be the first to potentially run locally, according to the company's description of it as "a version of 365 Copilot that is always working, able to take actions at any time."
OpenClaw's enterprise appeal
OpenClaw is an open-source tool that runs locally and creates agents to perform tasks autonomously. While it works on Windows, Mac Mini has become the preferred platform for OpenClaw users, driving unexpected sales growth for Apple's compact desktop.
The open-source project supports multiple models but Claude remains the most popular choice among users. Microsoft added Claude as an option for Cowork after partnering with Anthropic in late 2025.
Microsoft's enterprise-focused version would emphasize security controls that the open-source OpenClaw lacks, addressing a key concern for corporate deployments.
Expected announcement
Microsoft is expected to demonstrate the new agent or an upgraded version of existing tools at its Build conference in June 2026, according to The Verge.
Microsoft has not responded to requests for comment on how the new agent fits within its existing Copilot product lineup.
What this means
Microsoft is hedging its bets on AI agent architectures by pursuing both cloud-based and potentially local execution models. The focus on enterprise security and always-on operation suggests Microsoft sees a market gap between consumer-focused open-source tools and IT-approved business software. The proliferation of overlapping Copilot agents—Cowork, Tasks, and now this unnamed OpenClaw competitor—indicates the company is still determining which approach resonates with customers.
Related Articles
Anthropic offers K-12 teachers free year of Claude Pro with educational tools through June 2027
Anthropic launched Claude for Teachers, offering K-12 educators in the United States free access to premium Claude features for one year. The program includes Claude Cowork, Claude Code, and education-focused skills developed with Learning Commons, with applications open until June 30, 2027.
Microsoft Copilot adds PC Insights skill to analyze Windows settings and hardware
Microsoft is rolling out PC Insights, a new skill for the Copilot Windows app that allows users to query their system settings, hardware specifications, and software configurations. The feature accesses system resources with user permission but cannot make changes or automatically fix issues.
Anthropic adds sandboxed in-app browser to Claude Code desktop app
Anthropic has added an in-app browser to Claude Code's desktop application. The sandboxed browser allows Claude to read, click through, and interact with documentation, designs, and local development servers, with configurable session persistence.
GitHub reduces Copilot code review costs by switching to Unix-style exploration tools
GitHub reduced costs for Copilot code review by migrating to Unix-style code exploration tools. The company found that more sophisticated tools made reviews worse, leading them to reshape agent workflows around pull request evidence.
Comments
Loading...