Many organizations are dealing with a new, invisible tax on employees’ time: botsitting. This is the untracked, unrewarded work of repeatedly checking, correcting, and stabilizing an AI system before its output can be used. As a leader, you need to account for this labor to protect your team’s time and mental energy. Here’s where to start.
Identify when botsitting is necessary. Not all oversight is wasteful; some of it provides the accountability AI-assisted work requires, such as reviewing AI-generated material before it reaches a customer. Protect this work and build it into your team’s workflows.
Create clear metrics. Track how often AI-generated work needs human review, how many prompt attempts it takes to get a usable result, and how much time employees spend verifying answers or transferring context across tools.
Build psychological safety. If you reward token counts and login streaks, employees may quietly stop flagging bad outputs. Make it safe for people to say “this isn’t working” without fear of appearing resistant to adoption.
