AIThis post was created with the assistance of artificial intelligence (AI).

🔍 Read the full analysis: Can AI Reduce The Strain Of Switching Apps Throughout The Day? on ThorstenMeyerAI.com

Buying for a business?Offer from Amazon

Get business pricing on monitors, keyboards and dev gear

  • Business-only prices and quantity discounts
  • Tax-exempt purchasing
  • Multiple users, one account, clear invoices
As an affiliate, we earn on qualifying purchases.

TL;DR

A Harvard Business School preprint analyzing 103 million application events from 1,017 workers found that app switching was lower in the ten app uses following an AI use, but workers on AI-use days switched about 200 more times overall. The observational study cannot show whether AI caused either pattern, and it did not measure work quality or productivity.

A Harvard Business School preprint analyzing 103 million application events from 1,017 office workers finds that AI use was associated with both fewer app switches immediately afterward and more switching across AI-use days overall. The study, which is still under peer review and does not establish cause and effect, complicates claims that AI straightforwardly reduces the strain of moving between work apps.

Researchers Sumer S. Vaid and Ashley V. Whillans analyzed activity recorded from November 2025 through January 2026 at eight organizations in India. The data included 62,693 person-days. Workers switched between applications an average of 908.3 times per day; the median was 814. The study defines a switch as moving to a different application, not opening a different document within the same app.

The AI findings depended on the time frame. In the ten app uses after an AI use, workers used fewer applications, switched less, stayed roughly half a second longer in each app and followed more predictable sequences. But on days when workers used AI, they made about 200 more switches per day; days with more AI-use minutes were also more fragmented. These are associations in monitored activity, not evidence that AI caused more or less switching.

The researchers used data from Flowace, a commercial workforce-analytics service already used by the participating employers. Its software recorded active applications and related activity. Employers determined which apps counted as productive. The paper says data were anonymized before reaching the researchers and that Harvard’s human-subjects committee approved the protocol. The authors report that Flowace did not fund the research or take part in its design, analysis or interpretation.

At a glance
reportWhen: Preprint first posted July 7, 2026, rev…
The developmentA Harvard Business School preprint examining workplace app activity finds a mixed association between generative AI use and digital fragmentation.

App switching and AI · one moment

What happens around one AI query?

The study zoomed in on every AI use: the ten app uses before it, and the ten after. Up close, work calms down a little. Zoom out to the whole day, and the picture changes.

10 app uses before AI median 7 seconds 10 app uses after Whole AI day: about 200 more switches
  1. 1 · A workday of switchingWorkers changed apps 908 times a day on average. One app held their focus for 27.8 seconds on average, 3.3 seconds at the median.
  2. 2 · AI comes in at a busy momentPeople used AI a median of 13 times a day, for a median of 7 seconds. Fragmentation peaked in the hour of AI use.
  3. 3 · The next ten app usesFewer different apps, fewer switches, about half a second longer per app, more predictable. Real, but small.
  4. 4 · Zoom out: the whole dayOn days with AI use, the same people made about 200 more switches. Whether AI calms work or busy moments bring AI: unknown.

Source: Sumer S. Vaid, Ashley V. Whillans (Harvard Business School), “Digital Fragmentation and Generative AI Use Across 103 Million Application Events”, arXiv:2607.06681v2, preprint under review. Blocks are schematic, not to scale.

Graphic: Powered by Thorsten Meyer AI

Why AI May Not Mean Fewer Interruptions

The findings matter to employers and workers weighing AI tools as a way to make work less fragmented. The paper offers no basis for treating AI as a proven focus aid: it measured application activity, not the quality of work, output, stress or whether employees felt interrupted. A quieter sequence immediately after an AI interaction does not mean a less fragmented day overall.

The results also suggest that the unit of analysis matters. A single task sequence and a full workday can show different patterns. AI could be part of a day that is already more complex, or workers could use it in response to particular tasks; the study cannot distinguish those explanations. For organizations, the data are a reason to assess workflows and communication demands rather than assume buying an AI tool will reduce switching.

The researchers found that 44.6% of variation in fragmentation came from day-to-day changes within the same worker, compared with 35.8% from differences between workers and 19.6% from differences between organizations. That pattern points toward workload and scheduling as relevant factors, though it does not establish which specific practices improve focus.

Products Worth Considering

Logitech MX Mechanical Wireless Illuminated Keyboard, Tactile Quiet, Graphite

Logitech MX Mechanical Wireless Illuminated Keyboard, Tactile Quiet, Graphite

As an affiliate, we earn on qualifying purchases.

What the Monitoring Data Captured

The paper, titled Digital Fragmentation and Generative AI Use Across 103 Million Application Events, was written by Vaid of Harvard Business School’s Artificial Intelligence Institute and Whillans of its Negotiation, Organizations and Markets Unit. It was first posted to arXiv on July 7, 2026, and revised on August 7. The manuscript labels itself a preprint and says findings may change during peer review.

The sample covered eight Indian organizations in software, accounting, financial services, e-publishing, law and marketing. The researchers report a mean app-use duration of 27.8 seconds and a median of 3.3 seconds. Those measures describe observed time in an application before another app became active; they should not be read as direct measures of attention or uninterrupted thought.

The study also reports that fragmentation peaked on Tuesdays and reset after weekends and holidays. Its authors compare the app-use durations with earlier attention research, but note that the methods differ. The sample does not cover other countries or phone use, limiting how far the results can be generalized.

“The manuscript is currently under review, and the findings may change based on the peer review process.”

— Vaid and Whillans, in the preprint

Cause, Productivity and Wider Reach

The study is observational, not experimental. It cannot determine whether AI use led to more switching over a day, whether workers with fragmented days turned to AI more often, or whether both were shaped by the work being done. The short-term pattern after AI use does not resolve that question.

It is also unclear whether the measured switching harmed concentration or reduced output. The analysis does not establish whether workers completed tasks faster, produced better work or experienced less strain. The sample came from eight Indian organizations using one monitoring service, so it does not establish that the same patterns hold across countries, industries, remote-work settings or personal devices.

The preprint remains under peer review, and the authors warn that findings could change. The paper also discloses that Vaid holds equity in a private company unrelated to Flowace that uses AI to create enterprise digital twins; it says the research had no external funding.

Peer Review and Workplace Decisions

The immediate next step is peer review, which may lead to revisions or changes in how the results are presented. Until then, the reported numbers should be treated as findings from a preprint, not settled evidence that AI improves or worsens focus.

For organizations, the study supports checking when interruptions occur and whether communication or task scheduling contributes to them. Its authors suggest paying attention to concentrated morning work and midweek communication load, but this analysis did not test those practices. Any workplace changes should be evaluated against outcomes such as task completion and employee experience, not app-switch counts alone.

Further research would need to track a broader range of workers and devices, distinguish different kinds of AI tasks, and measure work outcomes. Experimental or longitudinal designs could help test whether AI itself changes switching behavior. The current paper does not provide that causal answer.

Key Questions

Did the study find that AI reduces app switching?

Not across the whole day. The researchers found fewer switches in the ten app uses after an AI use, but workers on AI-use days made about 200 more switches per day overall.

Does the research show that AI causes more switching?

No. The study is observational, so it identifies associations but cannot establish whether AI caused the higher daily switching count or whether other aspects of the workday explain it.

How many workers and organizations were included?

The analysis covered 1,017 employees at eight organizations in India, using activity records collected from November 2025 to January 2026.

Was the study peer-reviewed?

Not at the time described in the source material. It was posted as a preprint under peer review, and the authors say findings may change during that process.

Did the study measure productivity or work quality?

No. It measured application activity and switching patterns, not productivity, quality of work or employee stress. The results cannot show whether switching affected those outcomes.

Source: ThorstenMeyerAI.com

You May Also Like

Meta to sell excess AI computing capacity via cloud business, Bloomberg News reports

Meta plans to sell surplus AI computing power through its cloud business, Bloomberg reports, signaling a new revenue stream from its infrastructure.

Phone-Free Venue Teams: A Tool For Event Operations

An IdeaNavigator AI proposal outlines phone-free event operations software and recommends testing it at five venues and three schools.

OpenAI Divides Its Robotics Team While Runway Makes Praxis-1 Available

OpenAI has moved robotics staff into other groups, while Runway released an open-weight robot model built for faster action.

AI Job Losses And Return Offers: What The Evidence Shows

Surveys report AI-related rehires, but definitions and methods vary. Commonwealth Bank’s 45 job reversal is the clearest documented case.