AI was supposed to save office workers from drudgery, but it has handed them a shiny new pile of chores.
As artificial intelligence spreads through companies worldwide, it is removing some tedious old tasks while creating new ones. The Los Angeles Times reports that workers are spending plenty of time wrangling the tools that were meant to save them time.
UC Santa Barbara Duca Family professor of technology management Paul Leonardi said: “Most people don’t realise the amount of time that they’re spending working on the tools to get the time savings that they’re professing.”
Leonardi is one of the co-authors of a new study from the Work AI Institute, which includes academics from Stanford University and UC Berkeley. The institute is sponsored by AI outfit Glean.
The research surveyed 6,000 digital workers across the US, UK and Australia between December and January. It found workers are getting personal productivity gains, but companies are struggling to turn that into revenue or business growth.
While 75 per cent of individuals reported a productivity boost, only 13 per cent of organisations said AI adoption had delivered significant business gains. That is a fairly awkward gap for the corporate AI cheerleaders.
Leonardi said some of the productivity boost is wasted because workers must correct the bot’s output. They must gather the right files, documents and unspoken workplace knowledge needed to get decent results.
“It’s pretty striking the amount of time and effort people are spending,” Leonardi said.
Most employees now spend more than six hours a week babysitting their work chatbots, according to the survey. The report said there is a “thick, mostly invisible layer of human labour holding the whole thing together.”
For every hour a worker spends getting useful output from AI, they spend roughly another hour making it usable. Of the total time workers spend interacting with AI each week, 37 per cent goes to botsitting, while 36 per cent goes to producing actual work.
Part of the problem is that the tools still fall flat too often. Workers said more than a third of AI sessions fail outright, requiring a full restart or substantial rework.
As workers hand bigger chunks of their jobs to AI, they are offloading judgment and responsibility to the bots. The survey found 41 per cent of workers sometimes deliver AI-generated work that they could not explain if challenged.
Leonardi said: “We’re essentially expecting individual contributors to act as managers. They’re just managing these AI tools, AI agents, and we’re expecting that they’ll be able to produce way more, but we’re not taking into account all of the work that actually goes into managing.”
