TL;DR

New research from UC Berkeley suggests the workers most at risk of AI-related burnout are not the sceptics — they are the enthusiasts. After eight months studying a 200-person tech company, researchers found that employees who embraced AI voluntarily expanded their workloads until work filled every hour the tools freed up, and then kept going.

The productivity trap

The study, published in Harvard Business Review, tracked over 40 in-depth interviews with employees across a tech company where nobody was pressured to use AI and no new targets were set. Workers adopted AI tools on their own initiative and began doing more because the tools made more feel possible.

But the productivity gains came with a hidden cost. Work began bleeding into lunch breaks and late evenings. To-do lists expanded to fill the time AI saved — and then kept growing. Because the extra effort was voluntary and often framed as enjoyable experimentation, managers easily overlooked how much additional load workers were carrying.

The researchers concluded that “what looks like higher productivity in the short run can mask silent workload creep and growing cognitive strain.”

A cultural problem

The discussion around the research has highlighted a broader tension in work culture. AI tools make starting tasks easier and lower the barrier to taking on new responsibilities. But the result is not the promised reduction in work — it is an intensification of it.

The researchers warned that this voluntary intensity could impair judgement, increase errors, and mask unsustainable work patterns as genuine productivity gains. The risk is particularly acute because the overwork is self-imposed, making it harder for organisations to recognise and address.

Looking forward

As AI tools become standard across industries, companies will need to actively manage the human side of adoption. The UC Berkeley team recommends “intentional pauses,” deliberate sequencing of work, and protecting time for human connection. The core message: productivity without boundaries is not productivity — it is a path to burnout.