The use of AI in work is expected. Measuring it turns out to be much more complicated.
As American companies spend billions of dollars on AI software and agents, executives are ramping up efforts to measure adoption, monitor workflows and pressure employees to incorporate AI into their daily work.
From JPMorgan to Meta, companies have built dashboards and tools to track how their employees are using AI. Companies may also be collecting employee data needed to train their AI systems.
But as the use of AI becomes a workplace norm, new tensions are emerging.
While some employees are finding ways to inflate AI usage and games’ internal rankings, others are resisting and calling for stronger privacy protections.
The rise of AI dashboards
As the cost of AI increases, executives want evidence that their investments are being put to good use and that their employees are actually using the technology.
Many companies are turning to internal dashboards to track AI usage across teams and individual employees. Some of these dashboards are also published internally, allowing employees to see how well their colleagues are using AI.
For some employers, measuring AI adoption has become a way to identify laggards, compare employees, and influence decisions about performance.
JPMorgan, Meta, and KPMG are among the companies that have set up internal dashboards to monitor how much their employees are using technology.
However, tracking employees at this level of detail can backfire in some cases, as employees engage in “tokenmaxxing” (a strategy of using more AI tokens than necessary to manipulate metrics in their favor).
As the cost of AI rises, companies are becoming more concerned about this type of behavior, and some are rethinking how they use AI dashboards.
Amazon had an employee-created leaderboard that tracked the use of AI tokens, but shut it down in late May over concerns that it was encouraging employees to use AI unnecessarily to improve their rankings.
Surveillance in the age of AI
For years, workplace monitoring tools have focused primarily on productivity, tracking logins, mouse movements, screenshots, and time spent online, especially with the rise of remote work.
But AI is changing what companies want to monitor.
They observe how employees actually work, how they write, code, communicate, make decisions, and how they complete tasks.
This change is being driven in part by the rise of AI agents, software systems that can complete tasks with limited human oversight.
Capturing data about how employees work can help employers understand who is using AI effectively. It can also help companies train their own AI systems.
For example, Meta said in an internal memo in April that it would begin monitoring employee mouse movements and keystrokes so that the data collected could be used to train AI systems.
Promoting the introduction of AI
As tracking the use of AI becomes more common, some companies are using the data to make decisions about promotions, performance, and even job security.
Accenture CEO Julie Sweet said earlier this year that advancing within the company requires leveraging AI.
Some companies are trying to incentivize their employees by offering cash and other prizes.
For example, earlier this year, KPMG launched a program for its U.S. advisory sector that awards cash prizes to employees who use AI to generate new and innovative ideas for the business.