Published on the 18/02/2026 | Written by Heather Wright

Tools don’t reduce work, they consistently intensify it…

AI promised to give workers time back and free them up for the promised ‘value add’ work. It was to be a pathway to streamlined workflows and reduced operational burden. But new research shows AI is actually driving employees to work faster, take on more tasks and stretch their days further – often without being asked to do so.

While those early productivity surges might appear attractive to leaders keen for efficiency gains, academics from the University of California Berkeley’s Haas School of Business are warning that companies to take the blinkers off, saying they may not see what the gains are costing them until it’s too late.

“The productivity surge enjoyed at the beginning can give way to lower quality work, turnover and other problems.”

The eight-month study at a 200-person technology company examined how generative AI tools changed employees’ behaviours and workloads.

The researchers found consistent patterns. Employees broadened the scope of tasks they were willing to do, taking on responsibilities that previously belonged to other roles in what researchers say was a ‘meaningful widening of job scope’. Product managers and designers began writing code, researchers took on engineering tasks, and people took on work that would have normally been outsourced, deferred or avoided entirely.

“These actions rarely felt like doing more work, yet over time they produced a workday with fewer natural pauses and a more continuous involvement with work.”

And while they might not have felt like more work for those taking on the new, exciting tasks, they had knock-on effects that undoubtedly did create more work for others, including engineers who were forced to spend more time reviewing, correcting and guiding AI-generated or AI-assisted work produced by colleagues.

Employees also extended their working hours into additional parts of the day, often without being asked, the researchers report in Harvard Business Review. Typing into an AI system felt more like chatting than a formal task, making it easy for work to spill into evenings or early mornings and during what were meant to be breaks.

The third pattern was an increase in multitasking – widely regarded as increasing cognitive load – and increased expectations for speed, as people generated AI code while also manually writing code, ran multiple agents in parallel or revived deferred tasks for AI to handle in the background.

Sounds like a dream come true for business? The researchers warn that might not be the case, saying the changes brought about by enthusiastic AI adoption might be unsustainable. After an initial productivity surge, employees using AI reported more intense, busier, workdays and less work-life balance. Researchers warn cognitive fatigue, burnout and weakened decision-making can result.

“The productivity surge enjoyed at the beginning can give way to lower quality work, turnover and other problems,” they write.

Importantly, they note that the company didn’t mandate AI use – though it did offer enterprise subscriptions to commercially available tools.

Productivity questions

The results echo similar studies which have found that AI’s productivity outcomes vary widely, with CSIRO saying evidence is ‘murky’ about whether AI actually boosts productivity.

It’s study of 300 government organisation employees’ use of Microsoft Copilot found that while the majority self-reported productivity benefits, a 30 percent did not. And even those workers who reported productivity improvements expected greater benefits that were delivered.

A recent global Workday report also found employees were saving ‘meaningful time’ – 85 percent reported saving one to seven hours a week – with AI tools, but too often the gains were being absorbed by reworking low quality AI-generated content, creating a false sense of productivity and ROI.

The impact of workslop on co-workers has also been highlighted, as AI generated, ‘low effort’ work creates confusion and leaves co-workers attempting to decode content, infer missed or false context and potentially rework content.

Experts have also warned that AI as deployed by most organisations is providing only marginal improvements, at a personal level (though those personal gains are now called into question with the HBR report), rather than the reinventions required to drive true productivity growth across a business.

Building an AI Practice

The Berkeley researchers aren’t calling for a stop to AI. Instead, they make the point that AI will not deliver productivity gains unless organisations deliberately shape how it is use. Left unmanaged, AI increases speed, expands workloads and intensifies work in unsustainable ways – undermining the very benefits leaders hope to achieve.

To avoid this, the researchers argue companies much create a set of intentional norms and routines that structure how AI is used, when it is appropriate to stop and how work should. , and should not, expand in response to newfound capability.

They say asking employees to self-regulate is ineffective and businesses must instead define clear guidelines, in the form of an ‘AI Practice’ for when, why and how AI should be used.

That includes adopting intentional pauses, which they say would prevent ‘the quiet accumulation of workload that emerges when acceleration goes unchecked’ rather than slowing work overall. They cite the example of ‘decision pauses’ before major decisions are finalised, which would require a counterargument and an explicit link to organisational goals – widening the attention field enough to protect against drift.

The researchers also suggest organisations sequence work more deliberately, ensuring employees are not constantly multitasking across multiple AI-enabled workflows and add more ‘human grounding’ – short opportunities for coworkers to connect with each other through check-ins, structured dialogue or other means.

“Without intention, AI makes it easier to do more – but harder to stop. An AI practice offers a counter balance: A way to preserve moments for recovery and reflection even as work accelerates.”