As personal AI agents pick up steam, Microsoft is getting closer to launching its own in “Autopilot,” which is arriving first for business users, but will eventually launch for consumers, too.
As PCMag reports, the move is part of its revamp of the Copilot app, as Microsoft also touched on its effort to develop an AI agent that can work for enterprise customers around the clock, even while they’re asleep.
Microsoft first tipped Autopilots at its Build conference in June, describing them as “always-on agents that work autonomously, with their own identity, and act on your behalf.” At the time, Scout was introduced as Microsoft’s “first Autopilot agent,” but the company is now dropping that name and sticking with Autopilot.
Autopilot is Based on OpenClaw
Autopilot is based on OpenClaw, an open-source autonomous agent that can be installed locally on a PC to accomplish tasks. Microsoft likens it to a “digital teammate” that can work continuously without always requiring a user prompt, unlike AI chatbots. “Give it a name, a role, and a goal, and it goes to work—watching channels, following up on threads, running recurring work, and picking a project back up days later, without waiting for a prompt,” Microsoft says.
Microsoft released its AI agent project to a select group of enterprise customers in June at Microsoft Build. It’s now ready to expand it to more users, but still as a “private preview,” a sign that it’s moving closer to a wider release. In an interview with journalist Alex Heath, Microsoft’s CEO Satya Nadella also noted, “Autopilots will also go to the consumer side. In fact, we had an early version of all this, even earlier this year with Tasks.”
The Challenge is to Offer Autopilot without Causing Havoc
For now, a key difference is that Autopilot is designed to run from a cloud-hosted server, rather than locally on the user’s computer. Microsoft also plans to surface Autopilot through its apps. “It shows up where people already work—Teams, Outlook, chats, channels, and documents—so you can @mention it like a colleague, with permissions, audit, and governance behind it. You set the objective and boundaries; Autopilot handles the rest while keeping you informed and in control,” the company added.
The challenge is to offer Autopilot to users without causing havoc. An AI agent needs to harness the user’s computer, apps, or online accounts to fully reach its potential. But that access also means that if the AI agent makes a mistake, you could end up with data deleted, projects ruined, and your privacy exposed.
Building Guardrails to Make it Safe
Microsoft said in a briefing that it’s been building guardrails around Autopilot to make it safe for enterprise use. The company is betting that the technology can serve as a helpful worker for HR, sales, finance, and other business processes, allowing humans to focus on other tasks.
In a demo, Microsoft showed Autopilot acting as a digital worker to help a retail business keep its Black Friday sales on track by flagging overdue orders and ensuring all stores have enough inventory. The company sees the prospect of having “hundreds” of Autopilots working within a business’s IT environment to help solve problems in collaboration with human employees.
Microsoft CEO Satya Nadella, meanwhile, framed Autopilot as a key component to “building Copilot as a new OS for work.”
—
Photo Credit: Worawee Meepian / Shutterstock.com
