New Study of 443 Million Work Hours Reveals AI Is Making Work Denser, Not Lighter – And Meetings Are Part of the Problem

March 24, 2026 · 8 min read

The promise of AI in the workplace has always been simple: automate the tedious stuff so humans can focus on the important stuff. Less busywork. More thinking. Fewer hours grinding, more hours creating.

A massive new study suggests the opposite is happening.

On March 11, 2026, the ActivTrak Productivity Lab released its 2026 State of the Workplace report — one of the largest behavioral analyses of work activity ever conducted. The study analyzed over 443 million hours of digital workplace activity across 1,111 organizations and more than 163,000 employees, spanning 23 industries over a three-year period from 2023 to 2025.

The headline finding: AI isn’t reducing work. It’s compressing it. The workday is getting slightly shorter (down 2%), but collaboration has surged 34%, multitasking is up 12%, weekend work has increased more than 40%, and focus time has fallen to a three-year low.

In other words, AI is making every hour more intense, not making the day shorter. And meetings sit right at the center of this intensification.

The Key Findings

The ActivTrak data tells a nuanced story that defies the simplistic “AI replaces work” narrative.

AI adoption is surging. Eighty percent of employees now use AI tools at work — up 52% from two years ago. The average time spent in AI tools increased eightfold over the study period, and companies now use seven or more AI tools on average, up from just two in 2023.

Work is getting denser, not lighter. Despite widespread AI adoption, every category of work activity increased among AI users. Email volume is up 104%. Chat and messaging activity is up 145%. Business management tasks are up 94%. These aren’t small shifts — they represent a fundamental increase in the pace and volume of work flowing through each person’s day.

Focus time has hit a three-year low. The average focused session — an uninterrupted period of concentrated work — declined 9% among AI users, falling from 14 minutes and 23 seconds to just 13 minutes and 7 seconds. Meanwhile, focus time for non-AI users remained virtually unchanged. The tools designed to boost productivity are actually fragmenting the very deep work they’re supposed to enable.

Productive sessions are getting longer but fewer. While productive sessions grew 13% in duration (from about 24.5 minutes to 27.5 minutes daily), the total volume of focused time is still declining because the workday is more fragmented. Workers are starting earlier (7:48 AM vs. 8:02 AM) and working more weekends (up 40%), but the additional hours aren’t going to focused work — they’re going to the expanded coordination and communication that AI-accelerated workflows demand.

The “AI sweet spot” is narrow. Employees who spend 7-10% of their total work hours in AI tools show the highest productivity scores (95%) of any usage tier. But only 3% of users fall within that optimal range. Most are either underusing AI (and missing gains) or overusing it (and creating new forms of fragmentation).

What This Means for Meeting Costs

The ActivTrak findings have direct implications for how companies should think about meeting culture in the AI era.

If AI is making work denser — more communication, more collaboration, more coordination — then meetings are likely absorbing a growing share of the workday, even if individual meetings are getting shorter. The Cirrus Insight 2026 meeting statistics report confirms this pattern: meeting duration is declining on average, but meeting frequency is increasing, creating what they describe as “higher cognitive load rather than less.”

This means the cost-per-meeting may be dropping slightly (shorter meetings cost less in raw salary), but the total meeting burden is rising because there are more of them. A team that used to have three 60-minute meetings per week might now have five 35-minute meetings — spending roughly the same total time but absorbing five context-switching penalties instead of three.

The math on context switching makes this distinction critical. Research consistently shows that each meeting interruption costs 15-25 minutes of recovery time. Three meetings per week means 45-75 minutes of recovery time. Five meetings per week means 75-125 minutes. The difference — 30-50 additional minutes per week per person lost to context switching — adds up to 26-43 hours per year. At $41 per hour, that’s $1,066 to $1,763 in additional hidden cost per employee per year, just from the shift in meeting frequency.

At a 500-person company, this meeting density shift alone costs an additional $533,000 to $881,000 per year — and it’s entirely invisible unless you’re tracking it.

The Collaboration Surge Problem

The 34% surge in collaboration that ActivTrak measured isn’t inherently bad. Collaboration is how teams coordinate, align, and build things together. The problem is when collaboration time expands without a corresponding increase in the quality of what it produces.

More collaboration happening through meetings, Slack threads, email chains, and shared documents doesn’t automatically mean better coordination. It can also mean more noise, more duplication, more “keeping people in the loop” that doesn’t change anyone’s behavior, and more meeting-about-the-meeting cycles.

The ActivTrak data shows that AI users saw a 145% increase in chat and messaging activity. That’s not a small uptick — it’s a near-tripling of message volume. Some of that represents genuinely valuable async communication replacing unnecessary meetings. But much of it likely represents a new layer of coordination overhead layered on top of the existing meeting load rather than replacing it.

This is the trap many organizations are falling into: they adopt AI tools, which accelerate output, which creates more work products to review and coordinate around, which generates more meetings and messages to manage the increased volume, which fragments focus time further, which reduces the quality of the work being produced in those shrinking focus windows.

Breaking the Density Cycle

The ActivTrak data suggests that the solution isn’t less AI — it’s better communication design around AI-augmented work.

Calculate your true meeting load. Don’t just count meeting hours — count meeting frequency and the context-switching cost of each transition. A tool like Meeting Price Tag can show you the salary cost of each meeting in real time, but the hidden cost (recovery time, pre-meeting shadow, fragmented focus blocks) is typically 1.5 to 2 times the visible cost.

Target the 7-10% AI usage sweet spot. ActivTrak found that employees spending 7-10% of their time in AI tools had the highest productivity. For a typical 8-hour day, that’s roughly 34-48 minutes of AI tool usage. Beyond that, diminishing returns set in as the AI-generated output creates more coordination work than it saves.

Audit whether AI is creating or replacing meetings. For each AI tool your team uses, ask: is this tool reducing our need for synchronous meetings, or is it generating more work that requires meetings to coordinate? If AI is drafting more documents, does that mean more review meetings? If AI is generating more data, does that mean more analysis meetings? Track whether AI adoption is correlated with meeting reduction or meeting growth on your team.

Protect focus time more aggressively. With focus time at a three-year low, every minute of uninterrupted deep work is more valuable than ever. Meeting-free days, focus blocks, and async-first defaults aren’t just nice-to-haves — they’re essential counterweights to the collaboration density that AI is creating.

Make the cost visible. The ActivTrak data shows a clear “AI measurement gap” — most companies lack reliable data on how AI is actually changing productivity. The same is true for meeting costs. Making both visible — tracking AI’s impact on collaboration patterns and tracking the real-time cost of every meeting — gives leaders the data they need to make informed decisions rather than operating on assumptions.

The Bottom Line

The ActivTrak study of 443 million work hours delivers a message that every organization needs to hear: AI is not reducing the meeting problem. It’s reshaping it.

Work is getting faster, denser, and more fragmented. Collaboration is surging. Focus time is shrinking. And the workday is spilling into evenings and weekends as people try to find the quiet hours for the deep work that meetings are consuming during the day.

The organizations that will thrive aren’t the ones adopting AI fastest. They’re the ones that pair AI adoption with intentional meeting and communication design — ensuring that the productivity gains from AI aren’t immediately consumed by the coordination overhead that AI-accelerated work creates.