Artificial intelligence (AI) has rapidly transformed the way people work in offices. From preparing reports and analysing data to drafting emails and handling complex tasks, AI is helping employees complete work faster. However, a new challenge is emerging alongside these productivity gains. Employees increasingly have to review AI-generated work, identify errors and make corrections before the output can be used. This growing practice is being described as “Botsitting”, in which employees continuously monitor and correct work produced by AI systems.
According to a survey of around 6,000 digital workers in the United States, United Kingdom and Australia, nearly 87% of employees use AI in their daily work. Around 75% believe that AI has improved their productivity. The report indicates that employees save an average of 11 hours every week with AI. However, a significant portion of that saved time is spent checking and correcting AI-generated work. On average, 6.4 hours are reportedly spent reviewing AI outputs.
Work Has Become Faster, but Monitoring Pressure Has Increased
AI can generate content, organise data and complete several tasks within seconds. The problem arises when its output contains factual, technical or contextual errors. A machine cannot independently determine whether every fact, interpretation or conclusion it produces is accurate.
Employees therefore have to review important AI-generated outputs before using them. Constantly monitoring screens, identifying minor errors and correcting them can become mentally exhausting. Psychologists have pointed out that remaining continuously alert and checking details can contribute to increasing mental fatigue and workplace stress.
Detecting AI Errors Has Become a New Task
AI is often introduced as a way to save time, but the technology can also create an additional layer of responsibility. Employees first ask AI to complete a task and then have to assess the quality of the same work. If a significant error is found, they may have to rewrite or reconstruct the output.
This process can also encourage excessive dependence on AI. Employees who become tired of repeatedly reviewing machine-generated responses may eventually start accepting outputs without adequate verification. That can allow inaccurate information to move forward and potentially contribute to poor decisions.
Impact on Independent Thinking and Decision-Making
Growing dependence on AI could also affect employees’ ability to think independently and make decisions. If workers become accustomed to obtaining an initial solution from a machine for every problem, they may spend less time analysing issues themselves or considering alternative approaches.
Accountability remains another important concern when AI produces incorrect information. A machine cannot be held responsible for the consequences of a decision. Ultimately, the responsibility remains with the person or organisation using the output. This makes human oversight particularly important in sensitive or high-impact tasks.
Repeated Prompts Also Create Environmental Pressure
The impact of AI is not limited to employee time. Repeatedly asking questions, modifying prompts and generating multiple versions of the same response requires additional computing resources and energy.
Therefore, AI productivity cannot be measured only by the number of hours it appears to save. Organisations also need to consider the additional time employees spend reviewing outputs and the resources consumed when AI systems are repeatedly used to refine the same task.
Balance Between Human Judgment and AI Is Essential
To make AI an effective productivity tool in the workplace, organisations need clear guidelines governing its use. Instead of allowing machines to handle every task independently, companies should determine which activities can be automated and which require mandatory human verification.
Ultimately, AI’s speed can translate into genuine productivity only when it is combined with human judgment, decision-making and accountability. Without that balance, technology introduced to save time could become a hidden workload for employees, with workers spending more time correcting AI-generated mistakes than they originally saved by using the technology.
