Maximizing the generative artificial intelligence usage of employees has been a priority for many organizations. However, a number of concerns have called that into question, including the rising costs of using AI models, the need for AI quality control and oversight processes (typically conducted by humans), employee anxieties about job security, as well as challenges in measuring the return on investment (ROI) of AI usage.
These concerns beg a question for HR: should employee AI usage be a standalone performance metric, and if not, how might AI usage be effectively integrated into a performance management framework? SHRM spoke with industry leaders to get their perspective.
Reversals on Tokenmaxxing
Over the last year, we’ve seen reversals on tokenmaxxing, which is when employees maximize their AI token usage, and organizations encourage this behavior, often in the name of productivity.
For example, the learning-language platform Duolingo recently backtracked on its 2025 decision to make employee AI usage a standalone performance metric. Duolingo employees criticized the firm’s decision as driving AI for the sake of AI, rather than connecting usage of the technology to any specific business outcomes.
Speaking about the company’s reversal on a recent podcast, Duolingo CEO Luis von Ahn explained that “the most important thing in your performance is doing your job as well as possible. A lot of times, AI can help you do that, but if it can’t, we’re not going to force you [to use AI.]”
Another reversal in AI usage occurred at Accenture. The global consultant recently asked its own employees to stop using AI for simple and routine tasks like emails and presentations because of the firm’s skyrocketing AI usage costs.
The days of encouraging employees to use as much AI as possible seem to be changing. “Leadership should not be bulldozing their people into using AI... We’ve gotten a bit ahead of ourselves with pushing AI, and now we're seeing a backlash,” said Matt Poepsel, VP and godfather of talent optimization at The Predictive Index, a talent optimization platform.
Ongoing Concerns
While “cost discipline” around AI usage may be relatively recent, many employees have anxiety about the technology. SHRM's 2026 Q1 Global Employee Monitor found that 1 in 3 global workers are at least moderately concerned that their organization’s implementation of AI may lead to them losing their job in the next year.
In addition to fears of job loss, some employees have raised concerns about the low quality (and high error rates) of AI-generated output. The term “workslop” has become a common phrase to describe error-prone AI output, where AI-generated responses decrease work quality.
Finally, questions have emerged around how to measure the ROI of AI usage, and whether AI tools are delivering ROI at all. In a 2026 leadership survey of senior data and analytics leaders], 99% of respondents said investments in AI were a top organizational priority, yet fewer than one in five (18%) of these respondents say their organization is getting a high degree of measurable business value from AI. Additionally, a 2025 report from MIT found that only 5% of generative AI projects delivered any measurable return on investment.
5 Tips for Integrating AI Usage Within a Performance Management Framework
With reversals in tokenmaxxing and other ongoing concerns, how can organizations measure employee performance when it comes to AI usage? Some HR experts believe that AI usage should not be considered a standalone performance metric.
According to Dustin Snyder, founder and chief advisor at Wayforward Associates, a workforce consulting firm, AI usage doesn’t belong in the performance framework.
“A company should not be rewarding the use of an input without knowing whether it’s actually connected to an output. You give people an outcome and let them choose the method of achieving it,” Synder said.
But if an organization wanted to consider evaluating employee AI usage within their performance management, in some manner, how should it do so? Experts offered the following recommendations:
1. Communicate a good reason. “When AI is mentioned in our Glassdoor reviews, the number one thing people worry about is job security,” said Daniel Zhao, chief economist at Glassdoor.
“Employees need to understand why AI is being utilized and recognize the benefits they’ll get from using it. Even if your leadership team is incredibly excited about AI, you have to bring employees along with you.”
2. Connect AI usage to specific business objectives. “AI usage is a means to an end,” Poepsel said. “Start with your business objectives and then clearly describe what the impact would be if we used more AI.”
Every department will have their specific use cases and metrics, of course, and employees should be encouraged to experiment with AI to move the needle.
3. Evaluate and prioritize people (not tools). “We’ve got to change the performance management narrative to show how our people are becoming more effective,” Poepsel said. AI is best viewed like any other tool, as an instrument to improve human performance.
“In any performance management framework, we need to look at how the individual employee is valuable in a multi-dimensional way,” said Jackie Stinnett, vice president of people and great work at employee recognition platform O.C. Tanner.
Some questions Stinnett recommends leaders ask themselves to determine an employee’s value include:
- What is the employee’s expertise?
- How do they collaborate with other people?
- How do they add to our organizational culture?
- What is the employee’s level of productivity, and how does their use of AI help them achieve desired business outcomes?
4. Leverage competition. “Competitions are a great way to surface the best ideas for AI usage,” Zhao said. “Setting them up will require deciding what the focus should be, what the reward should be, and how you judge competitions. You might hold an AI hackathon where everybody at the company, not just the software engineers, can offer their best ideas for AI use that drives business metrics.”
A nice benefit of a competition-style program is that you can highlight the best ideas to the rest of the company via newsletters, reports, or videos. “If you just have a leader board that’s tracking employee AI usage, there's really not a lot of learning going on across the organization,” Zhao said.
5. Separate out metrics impacted by AI. According to Poepsel, organizations wanting to measure the impact of AI can do so by using a dashboard to share AI-backed metrics.
“That dashboard of ‘AI-influenced’ metrics offers the executive team a way to show the value that AI is helping deliver,” he said.
Move Fast, But Move Together
Cost discipline around AI and a more careful consideration of how employees use the technology is important. “It's bringing back some much-needed level-headedness around AI. The CEOs are right that we've got to move fast, but we’ve also got to bring people along with us,” Poepsel said.
HR leaders can lead their organization’s AI strategy by helping ensure AI is thoughtfully integrated into existing performance management frameworks.
Chuck Leddy is a Boston-based freelance writer who focuses on the impact of emerging technologies on the future of work, the modern workplace, and employee engagement. He's also written about and for institutions of higher education, including Harvard, MIT, and Boston University.
Was this resource helpful?
Validate your HR expertise
Earning your SHRM-CP credential makes you a recognized expert and leader in the HR field.
Related Content
Learn how Marsh McLennan successfully boosts staff well-being with digital tools, improving productivity and work satisfaction for more than 20,000 employees.
The proliferation of artificial intelligence in the workplace, and the ensuing expected increase in productivity and efficiency, could help usher in the four-day workweek, some experts predict.
As artificial intelligence technology continues to develop, the demand for workers with the ability to work alongside and manage AI systems will increase. This means that workers who are not able to adapt and learn these new skills will be left behind in the job market.