AI Use Was Not Linked to Work Exhaustion. Social Comparison Was.

An exhausted female worker sitting in front of a laptop with eyes closed and hand on forehead.
Pexels / Anna Tarazevich

A Finnish follow-up study of more than 2,000 workers found no clear link between frequent workplace AI use and exhaustion.

What stood out more was social comparison: people who were more inclined to compare themselves with others generally reported greater exhaustion.

A colleague finishes a task in minutes with AI. You are still figuring out which tool to use. The question can quickly shift from โ€œDoes this help me?โ€ to โ€œAm I falling behind?โ€

This kind of situation illustrates the social comparison that can arise at work. People with a stronger tendency to compare themselves with others reported more exhaustion. Workers who felt better prepared to understand and use AI generally reported less exhaustion.

Using AI often is not the same as feeling capable of using it. Someone can use AI every day and still feel uncertain about what they are doing.

Feeling prepared to use AI also mattered

The study measured perceived readiness rather than participantsโ€™ actual AI skills. It also did not test whether training would reduce exhaustion.

In one exploratory analysis, frequent AI use appeared to be associated with greater exhaustion among workers who compared themselves with others more strongly than average.

The primary analyses did not find this interaction between AI use and social comparison, so the subgroup result needs further study.

What the study leaves open

All the main measures were based on participantsโ€™ own reports. The exhaustion measure captured emotional exhaustion, one component of burnout, rather than the full syndrome.

The findings also do not establish which came first. The analyses examined associations across repeated surveys rather than showing that social comparison caused later exhaustion.

For employers, the study raises questions that go beyond AI adoption rates. Do workers feel capable of using the tools? What expectations surround them? Does learning happen collaboratively, or does it feel like a competition?

The study raises those questions but did not test solutions to them. A separate randomized trial among adults facing work and career difficulties found that a seven-week online ACT-based group program improved self-reported well-being but did not significantly increase career engagement. That trial did not investigate AI-related work pressures.

Iina Savolainen and colleagues at Tampere University followed employed adults aged 18 to 65 in Finland across three surveys in 2024 and 2025. The first survey included 2,109 participants, the second 1,680 and the final survey 1,479.

The study was published in SSM โ€“ Population Health on June 29, 2026.

RESEARCH REFERENCE

Reference: “Social comparison contributes to work exhaustion in the context of workplace AI use: A three-wave follow-up study of Finnish workers” by Iina Savolainen, Teijo Osma, Roope Grรถnroos, Moona Heiskari, Atte Oksanen, SSM โ€“ Population Health.
DOI: 10.1016/j.ssmph.2026.101945

Additional source: Tampere University Research Portal

Want to see more from HUMANUMi in Google?

More on This Topic

Latest News

More Latest News