Microsoft Trials Always-On AI Agents for 365 Copilot Inspired by OpenClaw

14 articles · Updated · TechCrunch · Apr 13

Microsoft is testing OpenClaw-like autonomous AI agents for its Microsoft 365 Copilot platform.
The company aims to create always-on agents capable of performing multi-step tasks for enterprise users, with improved security controls over open-source alternatives.
This move follows OpenClaw's popularity and reflects Microsoft's push to offer safer, proactive automation, likely to be showcased at its Build conference in June.

Can Microsoft truly secure 'always-on' autonomous Copilot agents against the deep risks seen in OpenClaw?

With other tech giants racing, what makes Microsoft's 'Frontier Suite' uniquely positioned to dominate the agentic AI market?

Could the rise of autonomous Copilot agents redefine office work, creating entirely new roles or displacing existing ones?

Beyond productivity, what unstated human or organizational costs might arise from widespread autonomous Copilot adoption?

How will 'always-on' Copilot agents balance continuous user monitoring with individual privacy and compliance standards?