🔍 Read the full analysis: Does AI Reduce Digital Distractions? The Harvard Study Weighs In on ThorstenMeyerAI.com
Get smart everyday buys 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 analyzed 103 million application events from 1,017 office workers in eight Indian organizations. It found fewer app switches in the ten uses after an AI interaction, but about 200 more switches on days when workers used AI; the observational data cannot show that AI caused either pattern.
The paper, Digital Fragmentation and Generative AI Use Across 103 Million Application Events, was written by Sumer S. Vaid and Ashley V. Whillans of Harvard Business School. It was first posted to arXiv on July 7, 2026, and revised August 7. The manuscript says it is under peer review and that its findings may change.
Researchers examined records from November 2025 through January 2026, collected by Flowace, a commercial workforce-analytics service used by the participating organizations. The data included 62,693 person-days. Across those days, workers made a mean of 908.3 app switches per day; the median was 814. A switch counted when the active application changed, not when a user opened a different document within the same app.
The apparent post-AI pattern applies to a narrow sequence: in the ten application uses after each AI use, workers used fewer apps, switched less and spent slightly longer in each app. The broader daily pattern went the other way: AI-use days had about 200 more switches, and days with more minutes of AI use were more fragmented. These are associations in monitored workplace data, not results from an experiment.
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’s Link to Workday Switching
The findings complicate claims that AI tools automatically reduce digital distraction. A worker might move through fewer apps immediately after asking an AI tool for help, yet still have a more fragmented day overall. The distinction matters for employers evaluating AI tools as a way to improve focus: this paper does not show that buying or using AI reduces interruptions.
The study also suggests that fragmentation may vary more with the workday than with a person’s or organization’s usual pattern. The authors attribute 44.6% of variation to day-to-day changes within the same worker, compared with 35.8% to differences between workers and 19.6% between organizations. That points toward workload and daily routines as relevant areas to examine, but it does not identify what caused the changes.
Products Worth Considering

Dell UltraSharp U2424H 24-Inch FHD Monitor

Dell 32 Plus 4K Monitor (S3225QS), 31.5-Inch VA, 120Hz, FreeSync Premium

FLEXISPOT 36-Inch Height-Adjustable Standing Desk Converter with Keyboard Tray
As an affiliate, we earn on qualifying purchases.
How Researchers Measured Switching
The sample included employees at eight Indian organizations in software, accounting, financial services, e-publishing, law and marketing. Flowace software recorded when the active application changed, along with information such as app name and window title. Employers determined which applications counted as “productive.” The research team received anonymized data under a data-use agreement, and the paper says the study protocol was approved by Harvard’s human-subjects committee.
The paper reports that a stay in one application lasted a mean of 27.8 seconds, while the median was 3.3 seconds. Those figures describe the study’s measurement of app use; they do not directly establish how long workers concentrated on a task. Comparisons the authors make with earlier attention research also involve different methods, so they should not be read as a precise measure of change over time.
The researchers say the work received no external funding. According to the paper, Flowace provided no commercial support and had no role in study design, analysis or interpretation. One author disclosed equity in an unrelated company that uses AI to create enterprise digital twins.
““The manuscript is currently under review, and the findings may change based on the peer review process.””
— Vaid and Whillans, in the preprint
What the Data Cannot Establish
The study cannot determine whether AI use caused workers to switch more or less. The authors say AI may reorganize work, workers may turn to AI on days when their work is already consolidating or fragmenting, or both patterns may be present. Because the records are observational, the direction of the relationship remains unknown.
The findings also cover a limited sample: employees at eight organizations in India, tracked through employer-used monitoring software. They do not establish what happens in other countries, on phones, or in workplaces outside the participating sectors. The study measured app activity, not the quality of work, productivity, stress or whether employees experienced the switching as distraction.
The headline claim that switching “drops after AI” describes the short sequence after an AI use, not the whole day. The paper’s different results at those two time scales should not be collapsed into a general claim that AI either improves or worsens focus.
Peer Review and Further Evidence
The paper remains under peer review, and the authors warn that its findings could change. Peer review may lead to revisions, but the study’s observational design will still limit what it can say about causation.
Stronger evidence about whether AI changes distraction would require research that can distinguish AI’s effect from the type and timing of work—for example, comparisons that track tasks and work outcomes alongside app use. For now, the preprint offers a large-scale description of switching patterns, not a test of AI as a focus intervention. Employers and workers should treat the results as preliminary evidence about app activity, rather than proof that AI improves concentration.
Key Questions
Did the Harvard study find that AI reduces digital distractions?
Not overall. It found fewer app switches in the ten application uses after an AI interaction, but about 200 more switches on days when workers used AI. The study does not establish that AI caused either pattern.
How many workers and app events were studied?
The researchers analyzed 103 million application events from 1,017 employees at eight organizations in India, recorded between November 2025 and January 2026.
What does the figure of 908 switches a day mean?
It is the mean number of changes from one active application to another per workday in the study. The median was 814 switches. Changing between documents in the same app did not count as a switch.
Has the paper been peer-reviewed?
No. The preprint was posted on arXiv and says it is under peer review. Its authors state that the findings may change.
Does the study show that AI improves productivity or concentration?
No. It measured application activity, not work quality, productivity or concentration. Its observational design also cannot show whether AI use caused the switching patterns.
Source: ThorstenMeyerAI.com
