'Botsitting' is destroying productivity as workers spend nearly a full day each week making AI 'usable'

The vast majority of UK digital workers (90%) now use AI in their daily activities, but productivity gains are being hampered as many fight to make the technology “usable”.

That’s according to a new study from Glean’s Work AI Institute, which found that AI appears to be having a positive impact on productivity. More than three-quarters (77%) said AI is making them more productive, helping to automate tasks and saving roughly 12 hours per week on average.

Yet these AI-related productivity gains are often offset by a new trend called “botsitting,” the study found.

Many workers are left “feeding it missing context, checking its outputs, debugging its mistakes, rerunning prompts, and cleaning up the confident-but-wrong answers AI leaves behind”.

Latest Videos From

Watch full video here:

All told, workers waste around 6.4 hours per week on average making AI usable, according to Glean, equivalent to nearly a full working day.

Notably, this is a problem that’s hitting UK workers harder than their international counterparts. The study, which polled digital workers in the US, UK, and Australia, showed that Brits spend a larger share of their time botsitting on average.

Cutting corners with AI

The long-term effect of this trend could have a significant impact on broader enterprise productivity, Glean warned. Frustrated workers cut corners with the technology to get projects over the line or simply file work that’s essentially slop.

“They stop checking outputs and deliver work they can’t fully explain or defend,” the report notes.

“That’s when botsitting turns into something more dangerous: botshitting – shipping AI-generated work that workers haven’t reviewed, don’t fully understand, or couldn’t defend if asked.”

Glean noted that more than two-thirds (69%) of workers polled admitted to ‘botshitting’ at work.

The AI productivity paradox

The research from Glean marks the latest in a string of studies warning about mismatched productivity gains with AI.

As ITPro reported in April, research from Accenture found workers are reporting higher levels of output and speeding up project delivery timelines with AI.

Yet these productivity boosts aren’t reflected on a business-wide level – and a key factor lies in the fact that underlying IT infrastructure or working practices haven’t evolved to compensate for the technology.

This same phenomenon was highlighted in the Glean study, with just 18% of respondents saying AI has “significantly improved” their organization’s overall performance.

Context matters

A sharpened focus on organizational change is critical for broader productivity gains, according to Glean – and shoehorning AI into the business is a sure fire way of scuppering plans.

Indeed, the study noted that those succeeding on this front typically “start with the work, selecting tools and platforms that fit the job instead of letting vendor contracts dictate their AI strategy”.

“Leading AI organizations resist AI addition sickness: the reflex to solve every problem by buying more AI, adding more tools, or pushing people to use AI whether or not it helps,” the study noted.

This tracks with a recent survey from Gartner, which found a leading factor in failed AI projects in integrating the technology in areas that aren’t applicable.

Similarly, these leading firms “understand that giving AI access to data is not the same as giving it context”. This is a major factor in why workers are spending so much time making the technology usable, the study warned.

More than half (53%) of workers said critical information they need to do their jobs isn’t accessible through their AI systems. In “context-poor” AI environments, workers are more than three-times more likely to feel worn out by AI.

By contrast, those in “context-rich” organisations are 52% less likely to ship AI slop work, spend less time botsitting, and are 31% less likely to “botshit”.

Rebecca Hinds, Head of the Work AI Institute at Glean, suggested that overcoming these challenges will require a rethink on how AI use is tracked in the enterprise.

“Too many companies are treating AI adoption like a vanity metric: more seats, more prompts, more usage,” she said. “But the UK data shows why that is not enough.

"British organizations have moved quickly to put structure around AI, and that is a real advantage,” Hinds added. “The next step is making sure AI is grounded in the right enterprise context, governed in the flow of work, and measured by whether it actually improves outcomes.”

FOLLOW US ON SOCIAL MEDIA

Follow ITPro on Google News and add us as a preferred source to keep tabs on all our latest news, analysis, views, and reviews.

You can also follow ITPro on LinkedIn, X, Facebook, and BlueSky.