🔍 Read the full analysis: What Researchers Say About AI And Workers' App-Switching Habits on ThorstenMeyerAI.com
Get hardware and tech essentials delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
TL;DR
A Harvard Business School preprint analyzing 103 million application events from 1,017 employees in India found an average of 908 app switches per workday. Workers switched less in the ten app uses after an AI interaction, but made about 200 more switches on days they used AI; the observational study cannot determine whether AI caused that pattern.
A Harvard Business School preprint analyzing 103 million application events from 1,017 office workers found that employees switched between apps an average of 908 times per workday. Its AI findings point in two directions: workers switched less in the ten app uses after an AI interaction, but on days they used AI they made about 200 more switches overall. The study is observational, so it cannot show whether AI use caused either pattern.
The paper, titled Digital Fragmentation and Generative AI Use Across 103 Million Application Events, is by Sumer S. Vaid and Ashley V. Whillans of Harvard Business School. Researchers examined 103 million application events recorded for employees at eight organisations in India from November 2025 through January 2026. The sample included workers in software, accounting, financial services, e-publishing, law and marketing.
Across 62,693 person-days, employees changed applications 56.9 million times. The mean was 908.3 switches per day, while the median employee made 814. The middle half of employees recorded between 547 and 1,130 switches per workday. The paper measured a mean app-use duration of 27.8 seconds before a switch; the median was 3.3 seconds. These figures describe application changes, not necessarily interruptions or lost productivity.
The AI analysis compared activity around uses of tools including Claude and other generative AI applications. In the ten app uses after an AI use, employees used fewer apps, switched less, remained in each app about half a second longer and moved between apps more predictably. Yet on days with any AI use, they recorded about 200 more switches per day, and days with more AI minutes were more fragmented. The authors say the direction of the relationship is unknown.
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.
- 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 · 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 · The next ten app usesFewer different apps, fewer switches, about half a second longer per app, more predictable. Real, but small.
- 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
AI Use and Daily Work Patterns
The findings complicate the idea that AI necessarily makes work more focused. The short-term pattern after an AI interaction is consistent with a more concentrated stretch of activity, but the broader daily comparison runs the other way: AI-use days involved more switching. Neither result establishes that AI improved or harmed work. The study measured app activity, not the quality, speed or value of completed tasks.
For workers and managers, the distinction matters when interpreting productivity dashboards or deciding whether to buy software as a focus aid. A day with more app changes could reflect fragmented attention, or a demanding workload that prompts both more switching and more AI use. The paper does not validate app-switch counts as a standalone measure of performance, and it does not show that introducing an AI tool will reduce interruptions.
The researchers also report that 44.6% of variation in fragmentation came from day-to-day differences within the same person, compared with 35.8% from differences between workers and 19.6% from differences between organisations. That distribution suggests work patterns varied considerably from day to day in this sample. The study describes fragmentation as peaking on Tuesdays and resetting after weekends and holidays, but it does not establish why.
Products Worth Considering

Apple MacBook Pro 14.2-inch with M5 Pro, 15-Core CPU, 16-Core GPU, 24GB Memory, 2TB SSD, Wi-Fi 7, Space Black

Apple MacBook Air 15-inch Laptop with M5 Chip, 16GB Memory, 512GB SSD, Midnight

LG 27GS60QC-B Ultragear 27-inch Curved Gaming Monitor (QHD, 180Hz, 1ms, 1000R)
As an affiliate, we earn on qualifying purchases.
How the Employee Data Was Collected
The records came from Flowace, a commercial workforce-analytics service used by all eight participating organisations. Its software records changes in the active application, including app names and timing, and may capture window titles and web addresses. Employers determined which apps counted as productive. The researchers received anonymised data under a data-use agreement, and Harvard’s human-subjects committee approved the protocol.
The paper states that it received no external funding and that Flowace provided no funding or commercial support and had no role in the study’s design, analysis or interpretation. It also discloses that Vaid holds equity in a private company unrelated to Flowace that uses AI to create enterprise digital twins. The authors are Harvard Business School researchers; the study was not conducted in a lab or as a controlled trial.
The manuscript was posted on arXiv on July 7, 2026, and revised on August 7. It identifies itself as a preprint under peer review and says its findings may change. The results therefore remain provisional. The sample covers a limited set of employers in India and does not include phone activity, workers in other countries or measures of whether work improved.
“AI may reorganise the day, people may turn to AI when their day is already consolidating, or both.”
— Sumer S. Vaid and Ashley V. Whillans, the paper’s authors
What the Switching Counts Cannot Show
The study does not establish that AI caused workers to switch apps more across a full day or less in the activity immediately after an AI use. Researchers observed existing work patterns rather than assigning workers to use AI under controlled conditions. It remains possible that employees turned to AI because their workday was already changing or unusually busy.
The measures also leave open what those switches meant for employees. App changes do not reveal whether a worker was distracted, collaborating, completing several tasks efficiently or producing better work. The paper does not assess output quality, task completion, stress or the experience of being monitored. Its results may not apply to workers outside the eight Indian organisations, to phone use or to different workplace monitoring systems.
The manuscript is a preprint under peer review, and the authors say the findings may change. Comparisons with earlier studies of how long people stay focused on a screen should also be treated cautiously because the methods differ. The reported durations are not a direct measurement of attention or concentration.
Peer Review and Workplace Evidence
The next formal milestone is peer review of the paper. The available material does not give a decision date or report a later journal publication. Until the review is complete, the findings should be read as provisional evidence about app activity in this particular sample, not as a general estimate for all workers.
Further research would need to test the relationship over time or compare similar workdays with and without AI use while accounting for workload and task type. It would also need measures of work quality and employee experience to determine whether switching patterns correspond to meaningful outcomes. For employers, the current data alone does not support treating AI as a proven focus tool or using raw app-switch totals as a direct score of worker productivity.
Key Questions
Did the study find that AI reduces app switching?
Not across the whole day. Workers switched less in the ten app uses after an AI interaction, but AI-use days had about 200 more switches overall. The study does not show that AI caused either pattern.
How many app switches did workers make?
The study reported a mean of 908.3 switches per workday across 62,693 person-days. The median employee made 814 switches per workday; the count refers to changes between applications, not necessarily interruptions.
Who and where did the researchers study?
The sample comprised 1,017 employees at eight organisations in India, across sectors including software, accounting, finance, publishing, law and marketing. Activity was recorded from November 2025 through January 2026.
Is the paper peer-reviewed?
No. The paper was posted as a preprint under peer review. Its authors state that the findings may change during the review process.
Does the study show that AI improves productivity?
No. It measured application activity, not task quality, output or productivity. The observed association between AI use and switching does not establish cause and effect or show whether workers performed better.
Source: ThorstenMeyerAI.com
