How People Use Copilot Depends On Device, Microsoft Reveals

Microsoft reveals that how people use Copilot varies by device, with different behaviors across desktop, mobile, and work environments.

Microsoft’s latest Copilot usage study suggests people treat the assistant very differently on their phones versus their desktops, and that the time of day strongly shapes what they ask. The report analyzes 37.5 million de‑identified consumer Copilot conversations from January to September 2025, using machine classifiers to group chats by topic and intent rather than human review.​

What The Report Finds

On mobile, “Health and Fitness” dominates across every hour and month, with users looking for both information and advice on their wellbeing.

The authors describe the pattern in the following manner:

“On mobile, health is the dominant topic, which is consistent across every hour and every month we observed, with users seeking not just information but also advice.”

Desktop activity follows a more work‑like rhythm. Overall, “Technology” leads as the top topic, but between roughly 8 a.m. and 5 p.m. “Work and Career” overtakes it on desktop, while education and science queries also rise during business hours.

Outside the workday, conversations skew more personal and reflective, with “Religion and Philosophy” climbing overnight, programming heavier on weekdays, gaming increasing at weekends, and a clear spike in relationship chats around Valentine’s Day.​

How Microsoft Interprets The Patterns

The authors argue that context especially device is a big driver of behavior. They write:

“This suggests a device-specific usage pattern where the phone serves as a constant confidant for physical well-being, regardless of the user’s schedule.”

In other words, smartphones are where people go for private, always‑on check‑ins, while desktops look more like a productivity surface that flips between work, learning, and evening curiosity.​

They summarise following broad “modes” of interaction on desktop:

• A workday mode focused on jobs.
• Tech and education.
• A “constant personal companion” mode more visible on mobile.
• An introspective night mode where philosophical and personal topics rise.

The intent mix also leans heavily toward “searching” and “advice” overall, reinforcing the sense that Copilot is being used both as an information tool and as a low‑friction way to ask for guidance.​

Important Caveats In The Methodology

Microsoft labels the analysis as a preprint, so it has not gone through peer review. It also covers consumer Copilot usage only, explicitly excluding enterprise‑authenticated traffic, which means it does not describe how Copilot is used inside Microsoft 365 or within workplace environments.​

All categorization comes from automated topic and intent classifiers, so the results reflect how Microsoft’s models group conversations rather than a human‑coded taxonomy. That is standard at this scale but worth remembering if you’re tempted to over‑interpret fine‑grained topic rankings.​

Why This Matters For Product And UX Thinking

The key takeaway is that “Copilot usage” is not a single behavior; it varies meaningfully by device and time of day. On phones, Copilot increasingly looks like a personal health and life‑management companion, while on desktop it leans into productivity, learning and late‑night reflection.​

For anyone designing AI experiences, this supports a more contextual approach: tailor prompts, UI, and safeguards to where and when people are using the assistant, rather than assuming a one‑size‑fits‑all pattern.

Bottom Line

Future research that includes enterprise usage, or cross‑tool comparisons beyond Microsoft’s own stack, would be valuable to confirm how broadly these patterns apply.