Study: AI does not make work easier, but more intense

Contents

Researchers at the Berkeley Haas School of Business warn: The use of Artificial Intelligence leads to employees working faster and sometimes even voluntarily extending their working hours – but all this possibly at the expense of their health. Fatigue, impaired decision-making ability, and burnout are threatened. As a kind of interim assessment of the study, which is still in progress, the researchers recommend that companies establish clear rules for the use of AI.

The eight-month field study, conducted between April and December 2025, examined a US-American tech company with around 200 employees. The company provided employees with access to commercial AI tools but did not mandate their use. The interim results of the study were published now in the Harvard Business Review.

Employees do more – and voluntarily

Employers may be pleased at first glance that the use of AI led to an intensification of work. Employees worked faster, took on a broader range of tasks, and extended their working hours – often voluntarily and without explicit instruction. The researchers suspect that AI gives employees the feeling that they can achieve more, and that this has a motivating effect.

They also observed changes in the work rhythm: fewer natural breaks were taken on workdays, and there was a constant switching between different tasks, facilitated by evaluating AI feedback and the growing number of tasks.

A partner in a conversational tone

Because interacting with AI in chat form feels more like chatting than formal work, and it is perceived as a partner, the boundaries between work and the rest of life have also blurred. Some work actions suddenly no longer felt like work, which is why the inhibition threshold for shifting work to the early morning hours and the evening also decreased.

However, the flip side is that after the initial enthusiasm subsides, a feeling of being overwhelmed sets in, as employees then suddenly feel the “silent increase” in workload. Health consequences include chronic fatigue and burnout. Companies risk increased employee turnover as a result. However, supervisors find it difficult to distinguish between genuine productivity gains and mere unsustainable intensity.

Recommendation: Set clear rules

To weaken or, ideally, prevent negative consequences, the researchers advise setting clear boundaries on how AI should be used and when its use should be stopped. Expansions of business scope should also be conscious decisions and not left to chance. Sustainable productivity gains can only be achieved if it is carefully determined how far AI should be integrated. It takes time and discipline to find a new balance.

(mki)