Artificial intelligence is delivering measurable productivity gains for office workers, with many employees reporting that AI saves them roughly 11 hours per week. However, a growing body of research suggests those gains are being offset by a new burden known as “botsitting”—the time spent supervising AI systems, correcting errors, providing missing context, rerunning prompts, and verifying outputs. Surveys of thousands of digital workers across the United States, United Kingdom, and Australia indicate that while individuals feel more productive, relatively few organizations are seeing corresponding improvements in overall business performance. The findings raise serious questions about whether the current AI boom is producing genuine efficiency gains or simply shifting work into less visible forms of oversight and quality control.
Sources
- https://www.latimes.com/business/story/2026-06-12/ai-saves-office-workers-hours-but-then-demands-hours-of-babysitting
- https://www.businessinsider.com/botsitting-ai-hidden-human-labor-at-work-2026-6
- https://www.itpro.com/technology/artificial-intelligence/botsitting-is-destroying-productivity-as-workers-spend-nearly-a-full-day-each-week-making-ai-usable
- https://www.ft.com/content/b4b60d00-2e8c-4db0-b3ed-9988dc0eeb5c
Key Takeaways
- AI is saving workers significant time on routine tasks, but employees are spending nearly a full workday each week monitoring, correcting, and refining AI-generated output, reducing the net productivity benefit.
- While individual workers report higher productivity, only a small percentage of organizations are experiencing substantial business-level gains, suggesting that AI adoption alone does not guarantee improved performance.
- Companies that successfully integrate AI appear to focus less on deploying more tools and more on improving workflows, providing context, training employees, and establishing accountability for AI-assisted work.
In-Depth
For years, the technology sector has promoted artificial intelligence as a revolutionary force that would eliminate repetitive work, free employees for higher-value tasks, and usher in a new era of productivity. While early results show that AI can indeed save substantial amounts of time, recent research suggests the reality is considerably more complicated.
The emerging phenomenon of “botsitting” highlights a problem that many executives seem reluctant to acknowledge. AI systems frequently require constant supervision, extensive prompting, correction of factual errors, and verification of results before their output can be trusted. In practical terms, many workers are finding that AI does not eliminate work so much as transform it into a different type of work. Instead of performing a task directly, employees increasingly spend their time managing a machine that performs the task imperfectly.
From a conservative perspective, this development serves as a useful reminder that technological hype often outpaces economic reality. Markets ultimately reward productivity, not promises. If workers are spending six or more hours each week correcting AI mistakes, then the true efficiency gains may be far smaller than headline claims suggest.
The more troubling finding is that businesses themselves are not seeing the dramatic performance improvements many expected. That suggests AI remains a tool rather than a substitute for human judgment, experience, and accountability. Companies that appear to be succeeding with AI are not simply replacing workers with algorithms; they are investing in training, workflow redesign, and human oversight.
The lesson is straightforward: AI may be a powerful productivity enhancer, but it is not a magic wand. Organizations that treat it as one risk discovering that the future of work still depends heavily on the people cleaning up after the machines.