While 90% of UK digital workers use AI at work and report personal productivity boosts, many find themselves dedicating substantial time to 'botsitting'—monitoring and fixing AI outputs—limiting broader organizational benefits.
- 90% of UK digital workers use AI, but only 42% in AI-first workplaces
- Workers save 12 hours weekly on average, with half spent monitoring AI
- 77% of workers recently corrected or redid AI-generated work
Market signal
AI adoption among UK digital workers continues to accelerate, with usage rates surpassing those in the US. Despite high engagement, the shift to AI-first organizational models remains limited, indicating a cautious or gradual integration approach. This mixed adoption landscape signals that while AI is becoming a standard workplace tool, full-scale transformation driven by AI is still in progress.
The data show a paradox where workers report significant productivity improvements individually, yet these gains do not translate into proportional organizational enhancements. This suggests that businesses must recalibrate expectations around AI-driven efficiency and consider the hidden costs related to error correction and output verification as inherent parts of current AI workflows.
Operator impact
Employees are increasingly tasked with 'botsitting'—spending time supervising, correcting, and redoing AI-generated work. On average, half of the potential time saved by AI automation is consumed by this oversight, fundamentally transforming job roles from task performers to quality controllers. This additional labor demand highlights the evolving nature of work where human operators remain essential to ensure AI reliability and accuracy.
The pervasive need for cleanup and correction of AI outputs poses operational challenges, including increased workload and potential employee fatigue. Organizations must recognize that AI tools currently require significant human supervision to be effective, complicating IT and management efforts to optimize AI return on investment.
What to watch next
As companies strive to derive more value from AI, a key focus will be developing better mechanisms for error detection, prompt crafting, and output verification. Metrics beyond simple time savings, such as the quality of AI output and the overhead of human intervention, will become critical for assessing AI effectiveness in the workplace.
The maturation of AI adoption could involve enhanced user training, improved AI reliability, and process redesign to reduce botsitting overhead. Monitoring how organizations balance AI productivity gains against supervisory burdens will provide insight into the evolving role of AI in business workflows and inform future technology investments.