For the past several years, artificial intelligence has been promoted as a workplace productivity game changer. Employees report that AI tools help them complete tasks faster, automate routine work, and reclaim valuable time during the workweek. In fact, a recent study by Glean’s Work AI Institute found that employees estimate AI saves them approximately 11 hours each week.

On the surface, that sounds like a remarkable return on investment. Yet the same research reveals that only 13% of employees believe AI has significantly improved their organization’s overall performance.

So where are those 11 saved hours actually going?

The Rise of “Botsitting”

The answer may lie in a growing workplace phenomenon researchers call “botsitting.”

Rather than spending all of their reclaimed time on higher-value work, many employees are investing substantial hours managing AI itself. This includes crafting prompts, supplying missing context, reviewing outputs, correcting mistakes, rerunning requests, and verifying information before it can be used.

According to the study, workers spend an average of 6.4 hours per week botsitting. That represents 37% of all time spent using AI tools, which is actually more than the amount of AI time devoted to productive output.

In other words, employees are often acting as supervisors for their AI assistants rather than being assisted by them.

The Information Problem

One major reason for this productivity drain is that AI is only as effective as the information it can access.

More than half of workers surveyed said the information they need is not readily available through their organization’s AI tools. As a result, employees spend time bridging gaps between systems, searching for context, and validating whether the AI-generated answer is accurate.

Organizations often assume that implementing an AI platform will automatically improve efficiency. However, if company policies are outdated, knowledge is fragmented across departments, or critical information exists only in employees’ heads, AI simply amplifies those underlying issues.

Without reliable and current organizational knowledge, employees are forced to spend additional time correcting or supplementing AI-generated responses.

Trust Matters as Much as Technology

Technology challenges are only part of the equation.

Employees also need confidence in how AI should be used and what level of oversight is expected. When workers are uncertain about accountability or fear being blamed for AI-generated errors, they tend to verify every output repeatedly.

This cautious approach is understandable, particularly in HR where compliance, legal risk, and employee relations issues demand accuracy. However, excessive verification can quickly consume the productivity gains AI was expected to create.

Clear guidelines, governance, and expectations help employees use AI more effectively while maintaining appropriate oversight.

A Hidden Retention Risk

The study uncovered another concern for employers: Employees who spend the most time managing AI are also more likely to be looking for a new job.

Workers who devote 40% or more of their AI time to botsitting were significantly more likely to be actively seeking other employment opportunities. They also reported higher levels of frustration and burnout related to AI use.

This finding suggests that organizations should look beyond AI adoption rates when evaluating success. High usage does not necessarily mean employees are having a positive experience. In some cases, it may indicate that workers are carrying an additional layer of invisible work simply to make the technology function effectively.

Training Makes the Difference

The organizations achieving the strongest results with AI are not necessarily those with the newest tools. Instead, they are the ones investing in employee development and process improvement.

The study found that employees in high-performing AI organizations were far more likely to report receiving adequate AI training and support. These organizations also formally recognize AI skills and view AI as an opportunity to redesign work processes rather than simply automate existing tasks.

Successful AI adoption is as much a people strategy as it is a technology strategy. Training employees on how to use AI effectively, establishing clear guidelines for verification, maintaining accurate organizational knowledge, and redesigning workflows can help ensure that AI delivers meaningful productivity gains.

What HR Should Measure

As organizations continue investing in AI, the most important metric may not be how often employees use the technology. Instead, employers should examine how much time workers spend managing AI versus being helped by it.

If employees are spending hours each week correcting outputs, searching for context, and validating responses, the promised productivity gains may never fully materialize.

AI undoubtedly has the potential to create efficiencies, but technology alone cannot deliver the results organizations are seeking. The real value emerges when employers combine effective tools with strong training, clear expectations, accurate information, and thoughtful workflow design.

Employers need to evaluate the perceived hours saved are truly being reinvested in productive work or quietly disappearing into the growing task of managing the technology itself.

Source: HR Daily Advisor