Microsoft is turning its flagship office software into components that run inside an AI assistant. On Tuesday, CEO Satya Nadella told attendees at the company’s Ignite event that Word, Excel and PowerPoint are now generally available inside Copilot, which he called “the new OS for work.”
The design flips the usual relationship between people and their software. For decades, a program like Word was the main attraction, with an assistant tucked in to help. Microsoft now envisions agents doing the driving, with the Office apps riding along as plugins. Those agents draw on an organizational knowledge layer that pulls context from emails, meetings, calendars, chats and files. Users can create documents, spreadsheets and presentations directly in Copilot, and the resulting files stay editable in their native Office applications.
The rollout is the latest step in a redesign of Copilot that Microsoft first announced in September. Nadella described parts of that vision as now becoming generally available, and autonomous agents and new governance tools are arriving alongside them.
What is Copilot Autopilot?
The most significant new piece is Copilot Autopilot, which Microsoft describes as a “persistent, proactive agent for personal productivity.” It is a cloud-based agent that users can hand responsibilities to, rather than single tasks. Autopilot can be authorized to work across services such as email and calendars and complete assigned tasks as part of its role. It will be billed by usage through Copilot Credits, but Microsoft has not announced final pricing ahead of general availability.
Microsoft also made generally available Microsoft IQ, an intelligence platform that includes Work IQ, a context layer designed to help agents understand a company’s communications, relationships and ongoing work. Nadella called Work IQ perhaps a company’s most valuable database. “It has a deep understanding of the people, their relationships to other people, your emails, your meetings, chats, projects that make up all the fabric of work,” he said.
That kind of access raises obvious questions. Whether companies want autonomous agents reading internal documents and messages will be a decision for each organization’s own IT review. Copilot Workspaces, a collaborative environment, is currently available through Microsoft’s Frontier early-access program. Microsoft says personal Autopilot agents will soon be able to join those workspaces. “Until now, AI interactions have mostly been single-player,” Nadella said. “Workspaces makes it a truly multiplayer experience, where people and agents drive towards a common goal, with shared context.”
Nadella returned repeatedly to trust, and he did not sugarcoat the problem. “We find ourselves in a place where alignment is elusive. Mechanistic understanding of these models is elusive. The question is: How can you even trust all this powerful technology? The reality is, by definition, you can’t,” he said.
How will Microsoft control autonomous AI agents?
His answer was not to make the models trustworthy. Instead, Nadella described “an architectural imperative to separate the supply of intelligence” from the authority granted to it. Microsoft says each Autopilot agent will have a unique ID, and administrators can restrict its permissions or block it entirely. All agent activity is logged for human review, and Autopilot can ask for human approval before taking certain actions.
Those controls are meaningful, but they also ask customers to rely on Microsoft’s own safeguards. How well those protections hold up against mistakes, rogue actions or data exposure remains to be shown in practice. Businesses will have to decide how much autonomy to hand over to the agents.
Microsoft also announced what it calls the “AI Transparency Initiative.” The effort has partnered with more than 40 major companies, including Adobe and Mastercard, to develop open, industrywide standards for AI governance and security. Nadella said the goal is to “help establish an open standard that makes AI safer, more transparent, and more accessible.”
On the cost side, Microsoft expanded Agent 365, its existing control plane for managing AI agents across enterprise environments. The update extends identity, security and management systems to AI agents, with what the company describes as “observability at every layer of the stack.” Microsoft says Agent 365 governs agent behavior in a way that is “autonomous in execution, but bounded and supervised.”
Does Microsoft track what AI agents cost?
A new piece, Agent 365 Business Analytics, is meant to show what AI actually costs and what it returns. “You can simply think of this as an ERP system of AI usage. You can see exactly what’s driving the spend and connect the spend to the outcomes that drive business value,” Nadella said. Customers will see real-world costs tied to AI token spending and can compare those figures with the value of the work performed. Microsoft acknowledged that its estimates of assisted work hours come from a model, not a direct measurement of time saved. Judson Althoff, CEO of Microsoft Commercial Business, made the sales pitch plainly: “anything you can do with Claude or ChatGPT, you can do better with Copilot, for less.”
Nadella spent little time on last week’s outage, which disrupted access to services including Exchange Online and Teams for users around the world. “Last week was a reminder of how critical digital infrastructure is as the dial tone for the world,” he said. “We know the trust you place in our services and take that responsibility extremely seriously, and in that context I know we have work to do.”
Microsoft also pointed to what it calls a “community-first AI infrastructure plan,” an apparent response to growing public concern over its data center projects. The keynote asked organizations to give software agents more autonomy at a time when Microsoft’s existing infrastructure is already under increased scrutiny.
This article was written with the assistance of AI.
News Factory APP - agentic news to boost your SEO & AEO.