Harvard study finds AI actually makes work harder rather than easier

A new Harvard Business Review case study indicates that generative AI tools might actually increase work for people, rather than lessen it – a surprising finding given the common claim that these tools will simplify tasks.

Researchers studied approximately 200 employees at an American tech company for eight months. They observed the employees at work and also conducted over 40 interviews with people from various departments.

Although generative AI helped employees work faster, studies showed it frequently led to more tasks overall, rather than fewer.

Study says AI increases expectations

A recent Harvard Business Review article explains that generative AI is changing work by making tasks faster and increasing the demand for more results.

Studies show that while AI tools help employees get more done, managers and teams often start expecting them to work faster and handle more tasks. Surprisingly, AI doesn’t usually reduce workload; it often leads to new duties and more work overall.

The study said that the widespread use of generative AI tools like ChatGPT actually intensified work.

The study found that as AI tools improved productivity, employees didn’t experience a decrease in their workload – instead, they faced increased pressure to do more. Workers reported needing to work faster and handle a broader set of tasks after AI was introduced into their routines.

The study found that AI helped workers switch between tasks more easily, leading to an increase in the amount of work completed.

The study showed that although AI helped employees do some tasks faster, those benefits were frequently cancelled out by the fact that it also led to more work in different areas.

The article points out that companies using generative AI might find it hard to tell if they’re actually becoming more productive, or just working at an unsustainable pace. Researchers call this phenomenon ‘unsustainable intensity’.

The case study is still ongoing.

Read More

2026-03-02 16:20