Digital debt—information piling up in inboxes and meetings faster than workers can handle it—has become a serious problem. Digital debt slows individual productivity and impacts the corporate bottom line, as people spend more time on administrative tasks than producing value. Against this backdrop, most leaders (60%) say² their organizations struggle to generate the innovation they need to stay ahead.

Microsoft is perhaps in a unique position to measure digital debt. Aggregated usage data from its suite of collaboration and productivity tools reveal that workers spend more than half their time (57%)² on communication (email and meetings) rather than on creation (documents and spreadsheets).

Survey data reveals a way out²: 70% of workers say they’d gladly offload as much work as possible to AI. More than two-thirds (76%) say they’d have no hesitation using it for administrative work. Similar numbers say they’d use it to assist them with analytical work (79%) or creative work (73%). Even more say they’d welcome AI’s assistance in getting the information they need to get their work done (86%) or summarizing meetings and action items from meetings (80%).

At the same time, business leaders’ attitudes toward AI should ease fears of an AI workplace takeover. Leaders said they were far more likely to turn to AI to boost employee productivity (31%) than to reduce headcount (16%).²

To maximize AI for productivity, organizations need to assess the best use cases for their business.

Gartner divides organizations into three broad categories³ when it comes to AI adoption:

• Productivity pursuers
• “Not in front of my customers” adopters
• More aggressive AI first/everywhere adopters

Identifying what category an organization belongs to helps it identify quick wins.

For example, productivity pursuers mainly want AI to help employees with administrative functions. These functions could encompass both the front and back offices, but productivity pursuers exercise caution when applying AI to mission-critical functions, such as research and development or supply chain. They also typically reserve AI for tasks that don’t involve direct customer interactions.

On the other hand, AI first/everywhere adopters are all in on AI and want to apply it as broadly as possible across the organization for back-office functions, core capabilities, and customer interactions. They may be comfortable relying on AI to enhance the customer experience, handle sales and marketing activities, and even customer service.

“With desired use cases in hand, company leaders should develop an organizational policy or set of guiding principles for AI,” advises Gartner.³ What are you, as an organization, comfortable doing with AI? What won’t you do? As Gartner says,⁴ “Without a clear definition of the lines you won’t cross, it will be impossible to know when you’ve crossed one.”

Among the features unveiled at the Microsoft event was Recall, which allows those who use Copilot+ PCs to scroll through snapshots of their work.

Recall lets users quickly find and see the context of videos and websites they’ve viewed, documents they’ve worked on, and more. They can even retrieve automated meeting transcripts and translations. The snapshots are encrypted and, true to the promise of embedded AI, processed on the PC instead of in the cloud, enhancing security as well as performance, according to Microsoft. Other Copilot+ PC features include AI-powered image editing and writing capabilities, again using embedded AI that saves time and enhances data security.

At the event, presenters also demonstrated real-time, AI-powered transcripts available for any video playing on one of the new laptops. It was a hint of more productivity-boosting enhancements to come thanks to AI.

Devices capable of AI workloads formerly available only on cloud services are also powerful enough to handle demanding tasks such as video editing or sophisticated number crunching with ease, promising even more value from AI-embedded devices.

We expect to see built-in AI capabilities as the standard for all PCs in the near term. Companies can take advantage of these additional features while transitioning to Windows 11 in advance of Microsoft’s last day of support for Windows 10 on October 14, 2025.