AI use at work may cause mental overload

A growing number of employees are using artificial intelligence in their work, with adoption highest among those with higher education, according to a recent survey by Statistics Finland and the European Commission.

In the February survey, 47 percent of respondents with a higher education degree reported using AI for work tasks. While AI is intended to make work easier and more efficient, its use can sometimes have the opposite effect, says Virpi Kalakoski, research director at the Finnish Institute of Occupational Health.

Kalakoski warns that AI use at work can strain mental resources and reduce well-being if workplace conditions are not properly managed. Employers must recognize that AI is changing work, she says, and productivity will not increase if workflows are not genuinely improved and employees end up feeling worse.

Overload from AI use can lead to what Kalakoski calls a “brain fry” — a term borrowed from translator Antero Tiittula’s Finnish rendering of the English phrase “AI brain fry.” This refers to a state of psychological and cognitive overload where one exceeds their capacity, leading to mental fatigue, difficulty concentrating, and a need for time off. Unlike burnout, which develops over time, brain fry is an immediate consequence of overloading, similar to muscle fatigue after intense physical exertion.

A March survey by Harvard Business Review of nearly 1,500 participants found that AI use increased workload, particularly in situations requiring multitasking, such as monitoring multiple AI agents and responding to their outputs. Kalakoski notes that multitasking is known to be highly taxing and inefficient for humans.

She describes this as a “decision load,” where humans must constantly evaluate AI-generated results. Since AI is not infallible and can make errors, employees must remain vigilant, and this continuous assessment is mentally demanding in a different way than producing work independently.

Trust in AI is essential for its workplace use, yet employees must simultaneously screen for potential errors. Kalakoski highlights that little discussion has addressed the new demand for critical reading skills that AI introduces. Beyond accuracy, workers must also assess whether AI-generated content is appropriate in tone and meaning.

While AI can handle routine and tedious tasks, freeing employees to focus on more demanding cognitive work, this shift can also make workdays more intense. Conversely, offloading dull tasks to AI can improve well-being by allowing workers to concentrate on their core skills, provided they feel in control of their work.

Kalakoski has interviewed workers across sectors who report pressure from employers to adopt AI quickly. Many employees seek clearer guidance on how and where to use AI in their roles, as generic examples offer little practical help. Unclear or conflicting instructions add to the strain.

Employers must communicate the purpose of AI and how it will improve work, Kalakoski says. Without this, employees may fear being replaced. She emphasizes that AI-related workload is a clear occupational safety issue, noting that Finnish workplace safety laws prohibit excessive strain.