New Delhi, Nov. 11 -- Employers that push staff to use AI all day may be encouraging the level of use that a large global survey links to weaker focus and lower productivity.
Sapien Labs surveyed 264,085 adults across more than 85 countries between June 2025 and April 2026. Of these, 176,198 used generative AI tools such as ChatGPT or Claude. People who used them several days a week reported the best mental wellbeing. People who used them several times a day or constantly reported the worst, scoring below even those who rarely used AI.
The study measured wellbeing with the Mind Health Quotient (MHQ), a 47-item assessment scored from -100 to +200. Sapien Labs puts the gap between moderate and constant use at around 20 points. Its earlier research, published in Population Health Metrics in 2024, found that a 12-point fall in MHQ corresponds to the loss of roughly one productive day a month. On that ratio, a 20-point gap equals somewhat more than a day and a half of lost output per person each month.
The same 12-point fall also corresponds to a doubling of the share of people in the distressed range, from 6.1% to 13%, the report says.
The workplace finding that matters most is about attention. Of the 27 mental capacities the MHQ measures, 22 followed the same curve as the overall score, rising with light use and falling with heavy use. Focus and concentration did not. It declined at every step up in use, from a few times a month through to constant use, and showed the sharpest fall of the capacities measured. Among the problems measured, confusion and slowed thinking showed one of the steepest rises.
In the 18 to 34 age group, which supplies a large part of most workforces, the Cognition dimension of the MHQ fell 21.2 points between people who used AI several days a week and those who used it constantly. That was the largest fall of the six dimensions measured.
The curve cannot be put down to people using chatbots for comfort or companionship. Among 18 to 34 year olds, those who used AI only for cognitive tasks such as writing, coding, analysis and planning showed the same rise and fall. Their average MHQ was about 61 at several days a week and about 43 among constant users.
Dr Tara Thiagarajan, founder and chief scientist of Sapien Labs, said work use carries its own risks. "Focus and concentration were worse even with moderate use, regardless of whether people used AI for cognitive or emotional purposes, and work is where most people do their most sustained thinking," she said. "If that thinking is offloaded to AI every day, the capacities people's jobs depend on may weaken, since the brain works on the same use-it-or-lose-it principle as the body."
She said people who use AI heavily because their jobs demand it will help answer the question of cause and effect. "People who use AI daily for their jobs have not chosen that level of use, and so their outcomes are less likely to simply reflect distress they already had," she said. "If their mind health declines as well, it strengthens the case that daily use itself is the cause."
The report is cross-sectional and does not prove that heavy use causes harm. Its authors lean towards a two-way relationship, in which poorer mental health draws people into heavier use and heavier use then worsens attention and compulsion problems. They note, though, that the push towards regular AI use at work makes it unlikely that heavy users are only people who were already struggling. They also found that poor sleep, diet and exercise were no more common among daily users.
Earlier research points the same way. A 2025 study of knowledge workers, presented at the CHI conference on human factors in computing, linked frequent AI use for cognitive tasks to reduced critical thinking. A 2025 MIT Media Lab study of essay writers found lower neural engagement among those who wrote with an AI assistant.
Thiagarajan said employers will have to respond. "Employers will need to take this into account, but until they do, those of us who use AI for work need to watch whether we are still doing the difficult thinking ourselves," she said.
For companies, the data puts the weight on how often staff use AI and which thinking it replaces. Sapien Labs says it will track the effects of work use and report on them.
Published by HT Digital Content Services with permission from PNN.