Change is good, right?
New studies about AI adoption and employee experience show the ups and downs of change—and why workers feel empowered as long as they have the proper supports
Key takeaways
A new report from Qualtrics looks at the impacts of AI adoption on workers and finds that some change is empowering
But as AI is allowing employees to do more, it’s leading to exhaustion
Companies need to take a people-first approach to AI adoption, rather than organization-first approach
When researchers at the UC Berkeley Haas School of Business set out to measure what happens when employees get access to AI tools, they expected to find relief. Instead, they found exhaustion. AI was speeding things up so people could work faster, take on a broader scope, and extend their hours. It made doing more feel possible, but it was burning employees out.
The issue isn't the technology, argues a new report from Qualtrics, the experience management company, it’s how companies are handling change management.
"Change doesn't appear to be the enemy here," says Benjamin Granger, PhD, chief workplace psychologist at Qualtrics. "Humans are exceptionally well-suited to adapt—that's what our brains are designed to do. What we conflate is the discomfort we feel at the prospect of change with an inability to handle it."
In their most recent study of the workplace, 2026 Employee Experience Trends, Qualtrics surveyed nearly 34,000 employees across 24 countries and found that employees who experienced significant change were actually more engaged and more productive than those who hadn't. But not all change is equal, and when employees try to manage the change on their own it can lead to “shadow AI” and a declining enthusiasm for AI. Companies getting adoption right are bringing a people-first approach, rather than top-down one.
Too little pressure isn't good for productivity, and too much isn't either. Humans actually prefer to work under a certain amount of pressure when they're supported. It's about striking a balance.
Benjamin Granger
Chief Workplace Psychologist, Qualtrics
Understanding AI change
Some change is good. Employees who went through significant change in the past year—new tools, new workflows, new ways of working—were more engaged and felt more productive than those whose work stayed largely the same, according to the report.
"Too little pressure isn't good for productivity, and too much isn't either," he says. "Humans actually prefer to work under a certain amount of pressure when they're supported. It's about striking a balance." Granger points to the Yerkes-Dodson Law, in which performance peaks under moderate stimulation, not in its absence.
What tips the balance is the source. With technology changes, process improvements, and strategic shifts, Qualtrics’ survey mostly found higher engagement across the board. But layoffs, reorgs, and leadership churn landed differently: Employees who went through those were far more likely to check out.
But even energizing change can tip into exhaustion when it's unmanaged. Qualtrics found that employees experiencing extreme amounts of change or pressure, of any kind, showed worse outcomes, such as more disengagement and lower well-being, just like those experiencing too little. As the Haas researchers found, many workers are trying to navigate through AI largely on their own. If the organization hasn’t provided the right tools or support to do all the work, people reach for others.
How employees are coping
The fundamental disconnect is that executives and employees aren't looking at AI the same way, and that's putting pressure on workers to bridge the gap themselves. Among those under high productivity pressure, 37% are sourcing their own AI tools outside sanctioned channels, compared to 23% under lower pressure.
The way to look at this is that shadow AI is a signal, not a compliance problem, notes Granger. When employees route around sanctioned systems, organizations lose visibility into which capabilities their people actually need and the opportunity to create governed pathways that give employees flexibility while keeping the organization in control.
Part of the problem, Granger says, is that most companies aren't treating AI transformation the way they'd treat any other major change. "When you're going through major transformation, that's precisely the time you need open dialogue with your employees," he says. "You need to be talking. How are they using it? Is it working? We've seen some organizations do that, but not all of them."
Employers: Think enablement, not autonomy
Perhaps due to the real and internalized pressures of AI’s potential, employee attitudes toward AI are leveling off after years of rising sentiment, according to Qualtrics. Granger points to internal campaigns for AI adoption: Here's what it does for the business, now use it. Employees need to know what it means for them. Without it, companies may struggle with ROI for their AI projects.
Read also: Can 'friction maxxing' help your organization?
"That's where organizations are missing a trick," Granger says. "People view new technologies through a very personal lens: ‘What's in it for me?’ When that question goes unanswered, or when the implicit answer seems to be you'll be expected to produce more, enthusiasm stalls."
The fix isn't a better communications campaign. Employees who rated their company’s AI workplace strategies more positively were those given tools for particular purposes rather than powerful general-purpose AI with no guidance.
"One of the places where companies fall down is with these Swiss Army knife tools," says Granger. "’Here you go, use it.’ That's like being given a lot of autonomy with no enablement."
The solution? According to Granger—give employees space to experiment and the attitudes will evolve on their own.
