Copy article

New Study Reveals Study Confirms AI Doesn't Reduce Work It Intensifies It

ended 11. February 2026

A new study just proved what burnt-out business owners already knew: AI doesn't free you up. It eats your day alive. An eight-month review of of 200 workers in a tech business found AI tools didn't reduce workload. They intensified it. People worked faster, absorbed tasks that weren't theirs, prompted during lunch, and blurred every boundary between "on" and "off" voluntarily. They felt 'more productive' but not 'less busy'.

Researchers want organisations to build "AI practice" rules around when to stop. This is fine if you have an HR department, but what about the solopreneur at 11pm, still prompting because it "only takes a second"?

The study found three problem patterns.

Task creep: AI makes everything feel doable, so you absorb three roles. Fix: a weekly "not my job" audit a colleague would have picked up. Bosses need to monitor this "helpful stepping in" and make a rule that the work should be passed over to the person who should normally be responsible for it. That could mean hiring an additional person in that role to maintain the benefits of the productivity increases for other colleagues collaborating with AI successfully.

Boundary collapse: Prompting during lunch, one last query before bed. AI has no idle state, so humans need to make sure they have one. Fix: a hard shutdown ritual. Close the browser. The tab will be there tomorrow ready to carry on.

Compulsive multitasking: Heavy AI users tend to have multiple AI threads running concurrently, and work on one task while another one is completed by AI. While this may feel productive, it causes significant cognitive debt flip-flopping between tasks and juggling all their current statuses. Batch AI work into focused blocks on one workflow, instead of letting it bleed across all your tasks all day.

We'd like your views:

  • Should workplace AI tools have enforced recovery time prompts, or should the user have the final say?
  • Are we repeating the social media attention trap, rebranded as productivity?
  • What does sustainable AI use look like for solopreneurs and small businesses with no HR safety net?
  • Whose fault is it when AI tools drive mental burnout by design?
  • Is "do more with less effort" the most dangerously hyped AI sentence in business right now?

2 responses from the Newspage community

Copy all

Star Quote
Copy

Colette Mason
AI Ethics Consultant at Clever Clogs AI
We've been here before. Social media was frictionless, borderless, impossible to put down. It took a string of mental health crises before anyone asked if that was deliberate. Now we're repeating it with AI productivity tools.

AI tools are designed with no cognitive recovery built in for the human. There's no "you've done enough." It's the social media playbook repackaged as productivity, and those most exposed are workers with nobody watching their back.

I've built enterprise systems for four decades. Every robust system has an idle state, where the machine waits for the human. Current AI tools don't pause, don't check in, don't care if you've been prompting since dawn. That's not a feature gap. It's a business model.

Harvard says the fix is organisational policy. That doesn't help the founder alone at midnight. What helps is AI with boundaries in the architecture. Until vendors treat cognitive recovery as a design standard, we're automating burnout and billing for the privilege.
Copy

Editor
Mitali Deypurkaystha
AI Governance Strategist & CMO at AISensa
“After 8 months and 200 workers, this study exposes the productivity paradox pressuring people, not freeing them. AI isn’t the root problem here, it’s an accelerant poured onto a culture that already worships hustle and punishes rest. When you’re a business owner there’s no clocking off bell, and AI just whispers that one more prompt at 11pm is proof you care, not a warning sign you’re cracking. We keep blaming individuals for burnout when the real culprit is a competitive system that applauds overwork and calls it ambition. Until we stop designing work around constant striving rather than human limits, AI will keep intensifying the damage, not delivering the promise.