Microsoft’s AI Copilot is getting a revamp aimed at making it faster, simpler, and more closely aligned with how customers actually use AI.
The company recently announced a series of internal shifts, naming Jon Friedman as Microsoft’s first chief design officer for Microsoft 365, the productivity suite formerly known as Office 365, and installing Jacob Andreou as EVP of Copilot.
The moves effectively merge the teams focused on business and consumer Copilot products. Part of the new leadership’s mandate is ensuring that Copilot, whether accessed through standalone apps, the web, or embedded across Microsoft software, actually meets users’ needs.
“How we move forward is to bend the speed of technology to the speed and need of humanity,” Friedman tells Fast Company.
Earlier versions of Copilot could feel cluttered, crowded with links and buttons pointing users toward a wide range of use cases. The redesigned experience pares that back considerably. Designers began with the prompt box itself, then layered in only the features users most commonly need, including options to start a new chat, revisit prior conversations, choose an AI model, or monitor long-running AI tasks.
“We literally started with a blank page for both mobile and big screen, and then we layered up the experience one step at a time,” says Friedman.
As users interact with Copilot, responses will generally appear above the prompt field. The prompt box itself now expands as needed and includes richer formatting options, making it easier to submit more detailed instructions. Features to refine requests, like specifying parameters for a generated image, will surface only when relevant through a design approach known as progressive disclosure. According to Microsoft, the streamlined app also loads more than twice as fast.
Within productivity apps like Word and PowerPoint, suggested prompts will adapt to what a user is doing. Someone opening a blank document, for instance, will see different AI suggestions than someone reviewing a heavily edited shared file. Users will also have clearer controls over how the AI interacts with their work, including options to limit Copilot to chat rather than editing, or to focus on a specific section of a document or spreadsheet.
“Whenever I’ve worked with like five or six people on writing a document, it’s been really hard to constantly just rewrite the document with chat,” says Friedman. “It’s way better to be able to fine tune these different areas of a document that you’re working on.”
So far, testing shows users given access to the new AI experience in apps like Word, PowerPoint, and Outlook increases their use of Copilot, according to Microsoft data.
Copilot updates will be launched first for enterprise customers before expanding to the full array of business and consumer users, and the changes are informed in part by Microsoft sitting with employees at partner companies and observing in detail how they use the product. But in terms of design, Microsoft is aiming for a kind of elegant user experience that’s still more common in consumer-facing apps, even in an era where high-end tech is routinely branded with a “pro” moniker.
“The bar for enterprise software is now the consumer bar,” says Andreou. “It is an expectation that things that you use at work are as good, as reliable, as functional, as easy to use as all the products that you know and you love in your personal life.”
For Microsoft, the distinction between consumer and enterprise use cases may not be as wide as it seems, especially when it comes to productivity software. Consumers still come to Microsoft for tools like Word and Excel, Andreou says, even if they use them differently than they would in an office setting. Under the hood, the systems powering those experiences are increasingly similar too. Pulling information from Bing for a consumer query, for example, is not radically different from retrieving answers from a company’s internal data stores.
“Productivity is really a whole life thing, and it really is the scenario that changes, but the tool set is very powerful across whole life,” he says.
Both consumers and business users have no shortage of AI tools to choose from, including apps from AI labs like OpenAI and Anthropic and AI assistants integrated with everything from business software like Slack or Asana to consumer-centric tools like web browsers and smartphone interfaces.
Microsoft itself has faced some criticism for rapidly deploying AI features in a wide array of products, from longtime Windows utilities like Paint and Notepad to its various enterprise tools, leaving some users bewildered about the seemingly massive array of Copilots and what each can and can’t do.
But arguments for choosing Microsoft’s software over competitors have always included its ubiquity and the integrations between its products. And those now include AI-focused features like Work IQ, which can let Copilot securely access an array of company data stored in Microsoft 365 software, whether the AI is invoked from within apps like Word or Excel.
In the future, says Andreou, Microsoft may even be able to roll out greater AI-powered integrations between the consumer and business side of things, perhaps enabling people to do tasks like plan a vacation based on information from both personal and work calendars.
Such changes won’t come immediately, though, since the company will need to account for human expectations and needs around security and privacy at least as much as the raw capabilities of the AI.
“All of those things rooted in that brand promise of compliance and security have to very much continue,” Andreou says. “And so that’s kind of the path that we’re charting as we look to the future, bringing these two things together.”
