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

🔍 Read the full analysis: Could AI Help Cut The Interruptions Behind App Switching? on ThorstenMeyerAI.com

Before you orderOffer from Amazon

Get the latest gadgets delivered free with Prime

  • Fast, free delivery on millions of items
  • Prime Video, Amazon Music and more included
  • Member-only deals all year
Start your free Prime trial Free trial for eligible customers · Cancel anytime
As an affiliate, we earn on qualifying purchases.

TL;DR

A Harvard Business School preprint analysing 103 million application events found workers switched apps an average of 908 times per day. App use was more orderly in the ten uses after an AI interaction, but workers made about 200 more switches on days they used AI; the observational data cannot establish why.

A Harvard Business School preprint analysing 103 million application events from 1,017 office workers found that employees switched apps an average of 908 times per workday—but it does not show that AI reduces those interruptions. Workers made fewer switches in the ten application uses after an AI interaction, while days on which they used AI involved about 200 more switches overall, a contrast that matters to employers considering AI as a way to improve focus.

The study, titled 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. The researchers analysed second-by-second application records from 1,017 employees at eight organisations in India, covering November 2025 through January 2026. The paper was posted on arXiv in July and revised in August; it says it is a preprint under peer review and that its findings may change.

Across 62,693 person-days, employees made 56.9 million app switches. The mean was 908.3 switches per day; the median worker made 814, and the middle half of employees ranged from 547 to 1,130. These figures count changes to a different application, not movement between documents within the same app. The study also reports a mean app-use interval of 27.8 seconds and a median of 3.3 seconds before the next switch.

The AI results differ by time scale. In the ten app uses following an AI use, workers used fewer apps, switched less and stayed about half a second longer per app. But on days when workers used AI, they made about 200 more switches, and days with more AI minutes were more fragmented. The authors did not test whether AI caused either pattern.

At a glance
reportWhen: Preprint posted July 7, 2026, revised A…
The developmentA large Harvard Business School preprint reports that app switching was lower immediately after AI use but higher overall on days workers used AI.

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

AI Use and the Focus Trade-Off

The findings caution against treating AI as a proven focus or productivity tool. A short run of less fragmented activity after using an AI application does not mean the whole workday becomes calmer: the daily association in this sample went in the opposite direction. Employers weighing AI tools should distinguish between a local pattern following a particular interaction and an overall change across the day.

The study also suggests that interruptions may depend substantially on the day’s conditions, rather than being a fixed trait of a worker or company. The authors attribute 44.6% of variation in fragmentation to day-to-day changes within the same person, compared with 35.8% to differences between people and 19.6% to differences between organisations. That makes scheduling and communication load relevant questions, while leaving open whether changing them would improve work or outcomes.

For workers, the results are descriptive rather than prescriptive. The paper does not establish that app switching harms performance, nor that reducing it would improve the quality of work. It measured digital activity, not whether employees completed tasks faster, made fewer errors, or felt more focused.

Products Worth Considering

Logitech MX Mechanical Mini for Mac Wireless Keyboard, Space Gray

Logitech MX Mechanical Mini for Mac Wireless Keyboard, Space Gray

As an affiliate, we earn on qualifying purchases.

How the Workday Data Were Collected

The records came from Flowace, a commercial workforce-analytics service used by the eight participating organisations. Its software recorded when the active application changed, along with information such as app name and start and end times. The employers decided which apps counted as productive. The researchers received anonymised records under a data-use agreement, and the paper says Harvard’s human-subjects committee approved the protocol.

The sample covered software, accounting, financial services, e-publishing, law and marketing organisations in India. It did not include phone use or workers in other countries. The paper says it received no external research funding and that Flowace had no role in study design, analysis or interpretation. One author disclosed equity in a separate private company that uses AI for enterprise digital twins.

The paper compares its brief app-use intervals with earlier attention research, but cautions that the methods differ. Those figures should be read as a possible trend signal, not a precise like-for-like measure of how attention has changed over time.

“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 Switching Data Cannot Show

The study is observational, not an experiment. It cannot determine whether AI use led to more switching, whether workers used AI because their workday was already fragmented or consolidating, or whether another factor influenced both. The lower switching immediately after AI use and higher switching across AI-use days are associations at different time scales, not proof of a contradiction or a causal effect.

It is also not yet clear whether the patterns apply beyond the sampled organisations in India, or whether they hold for other kinds of work, phone use or AI products. The data measure application activity, not work quality, output or employee wellbeing. Peer review is ongoing, and the authors say the findings may change.

Peer Review and Broader Evidence

The immediate milestone is peer review of the preprint. Readers should check for a reviewed version or later revisions before treating the estimates as settled. Further research would need to test the direction of the relationship, include broader worker and device samples, and measure outcomes such as task completion and work quality.

Until then, the paper supports a limited conclusion: app switching was common in this sample, and its relationship with AI varied depending on whether researchers looked at the next few app uses or the whole day. It does not show that purchasing an AI tool will reduce interruptions. The authors’ data also point to day-to-day variation as a major part of the pattern, but do not test whether protecting focused work periods changes it.

Key Questions

Did the study find that AI reduced app switching?

Not across the whole day. Workers switched less in the ten app uses after AI use, but made about 200 more switches on AI-use days overall. The study does not establish that AI caused either result.

How many times did workers switch apps?

The reported mean was 908.3 switches per workday; the median was 814. These are figures from the 1,017-worker sample, not a universal estimate for all office workers.

Was this a peer-reviewed study?

No. The paper was posted as a preprint under peer review. Its authors state that the findings may change during that process.

Can the study show that AI causes interruptions?

No. It analysed observed application records rather than assigning workers to use AI. The data cannot distinguish whether AI affects switching or whether workers’ work patterns influence when they use AI.

Who and what did the study cover?

It analysed records from 1,017 employees at eight organisations in India, collected from November 2025 to January 2026 through Flowace workforce-analytics software. The study did not cover phone use or measure work quality.

Source: ThorstenMeyerAI.com

You May Also Like

My News Pipeline’s AI Results: The Best Configuration Isn’t In Use

A 71-configuration test found GPT-6 Astra scored highest for news writing, while cheaper GPT-6 Luna remains in production and unsafe fallbacks were removed.

2026’S Most Advanced AI 4K Webcams: Top 9 Choices

Discover the nine best 4K webcams of 2026, featuring top choices like Logitech Brio and Acer A640, for streaming, professional use, and content creation.

Does Microsoft’s Offer To 60,000 Reflect A Broader Shift In AI At Work?

Microsoft’s voluntary exit offer covers about 60,000 employees. The company says it is not tied to AI; its uptake and effects remain unknown.

Data Centers Surges In Global Coverage

Data center coverage worldwide has spiked, with 94 mentions in recent monitoring, indicating rising global interest and potential industry shifts.