Workers are spending over 6 hours a week 'botsitting' AI, fueling job frustration (www.businessinsider.com)

A recent report from Glean's Work AI Institute reveals that white-collar workers are spending an average of 6.4 hours weekly "botsitting" AI systems—tasks that involve contextualizing data, debugging outputs, and correcting errors. Despite 87% of surveyed employees in the US, UK, and Australia claiming that AI enhances productivity, only a mere 13% reported noticeable improvements in organizational performance. The term "botsitting," introduced in this study, underscores the often-overlooked yet burdensome work required to make AI functional, leading to worker fatigue and low morale. The significant disconnect between potential productivity gains and actual company performance poses a productivity paradox for organizations. Employees engaged in extensive botsitting are 73% more likely to seek new jobs, which indicates that the added workload is diminishing job satisfaction. Researchers argue that merely deploying more AI isn't the solution; organizations seeing real improvements focus instead on supporting employees with adequate training, context-setting, and clear standards for effective AI usage. Failure to address these challenges risks not only ongoing dissatisfaction but also the loss of skilled workers weary of "cleaning up after the bots."