A study of 164,000 workers just proved AI makes you busier — not more productive

AI was supposed to save you hours every week. A massive new study says the opposite is happening — and the numbers are hard to ignore.

ActivTrak's 2026 State of the Workplace Report tracked 163,638 employees across 1,111 organizations over three years and 443 million hours of work. The conclusion: AI isn't freeing people up. It's loading them with more tasks, more messages, and less time to think.

The numbers behind the overload

Among the 10,584 workers tracked before and after adopting AI tools, every type of digital activity went up — and nothing went down:

In other words, AI helped people produce more output — and then their employers immediately filled that freed-up time with even more work.

Why 'workload creep' is the real AI danger

Gabriela Mauch, ActivTrak's chief customer officer, summed it up: "The capacity freed up immediately gets repurposed into doing other work, and that's where the creep is likely to happen."

Separate Harvard Business Review research confirmed the pattern. Researchers studying a tech company found that employees who voluntarily adopted AI started taking on unsustainable workloads without realizing it. AI raised their managers' expectations, which made them even more dependent on AI to keep up — a vicious cycle the researchers called "workload creep."

Aruna Ranganathan, a researcher at UC Berkeley, warned that while AI boosts short-term output, it "can lead to cognitive overload, burnout, poorer decision-making, and declining work quality."

Only 3% of workers are using AI right

The ActivTrak study found a narrow sweet spot: workers who spent 7–10% of their day using AI tools hit 95% productivity — the highest of any group. But only 3% of workers actually land in that range.

The majority — 57% — spend less than 1% of their work hours using AI. Meanwhile, heavy users face what researchers now call "AI brain fry": mental fatigue from juggling too many AI tools, constant task-switching, and supervising multiple automated processes at once.

The gap your boss doesn't see

The report's central finding is what ActivTrak calls the "AI Measurement Gap" — most organizations adopted AI quickly but have almost no visibility into whether it's actually helping. Collaboration surged 34% and multitasking rose 12%, but nobody's measuring whether any of that extra activity produces better outcomes.

There's a small silver lining: burnout risk actually fell 22% to just 5% of workers, and 75% maintain healthy work patterns. But disengagement — the quiet, creeping sense that the work never ends — is rising fast.

If you manage a team that uses AI

The study's practical takeaway is clear: don't measure AI success by how much work gets done. Measure it by what happens to the people doing the work.

Track whether your team's focus time is shrinking. Watch for "always-on" chat patterns. And resist the urge to fill every minute AI saves with another task — that's exactly how the creep starts.

Related Content — Get Started with Easy Claude Code | Free Learning Guides | More AI News