UK digital workers spend 6.4 hours weekly ‘botsitting’

A new study by Glean’s Work AI Institute found UK digital workers spend an average 6.4 hours a week ‘botsitting’—providing missing context, checking outputs, rerunning prompts and fixing errors-while reporting AI saves about 12 hours of work per week.

Glean surveyed digital workers in the UK, US and Australia. The report states 90% of UK respondents use AI daily and 77% said AI makes them more productive, citing automation of routine tasks and faster task completion. Respondents reported saving roughly 12 hours a week on average but also said they spend about 6.4 hours weekly debugging and cleaning up AI results to make the tools usable.

The study found UK workers spend a larger share of time botsitting than their international peers. Only 18% of UK respondents said AI has ‘significantly improved’ their organisation’s overall performance.

Glean identified a related behaviour it calls ‘botshitting’—submitting AI-generated work that employees have not fully checked or cannot explain. More than two-thirds (69%) of those polled admitted to doing this.

Access to data and context emerged as a key problem. Fifty-three percent of respondents said critical information needed for their jobs is not accessible through their AI systems. In workplaces described as ‘context-poor’, workers were more than three times as likely to report feeling worn out by AI. In ‘context-rich’ environments, employees were 52% less likely to deliver poor-quality AI outputs, spent less time botsitting and were 31% less likely to admit to botshitting.

The report linked certain organisational choices with better outcomes. Firms that choose tools to fit specific tasks rather than letting vendor contracts dictate strategy, and that avoid adding AI tools to every problem, reported better results. The study also states that measuring AI adoption by seats, number of prompts or raw usage does not guarantee improved outcomes.

Rebecca Hinds, head of the Work AI Institute at Glean, warned: ‘Too many companies are treating AI adoption like a vanity metric: more seats, more prompts, more usage.’ She added that British organisations have moved quickly to put structure around AI and that the next step is to ground AI in enterprise context, govern it in the flow of work and measure whether it improves outcomes.

Glean cautioned that without better access to context and clearer governance, the time spent keeping AI usable could erode the broader productivity gains organisations expect.