Mental Health Care Was Already a Priority for Organizations. AI Raised the Stakes.

More Chief People Officers are naming mental health as one of their top priorities. When you see names like JPMorgan Chase and Dell Technologies being singled out for excellence in mental health, you know this is a clear business priority. And with good reason.

Disability leaves are up sharply this year. Sixty-five percent of benefits leaders report drastic increases, and the number of people entering care with severe conditions has grown 46% since 2021, according to Lyra Health's 2026 workforce data. Complex conditions like severe depression and anxiety are up 88% year over year. Nearly 7 in 10 benefits leaders now say mental health challenges are significantly affecting employees' ability to do their jobs.

But why the spike? A survey of more than 1,100 U.S. employees by Mind Share Partners and Qualtrics found 90% reported at least minor levels of a mental health challenge, with the top three stressors being U.S. politics, global events, and personal finances. And it's happening at all levels of organizations. Fifty-five percent of CEOs reported a mental health issue of their own in the past year.

At the same time, more than a third of benefits leaders say AI specifically is driving new anxiety, and this unique stressor needs to be addressed concurrently.

The instinct, for the general spike, is to respond with more support. Better EAPs. Expanded mental health benefits. A wellness stipend. Companies are spending more on these programs than ever.

These are worth doing, but they aren't enough on their own. A wide body of workplace wellness research shows individual-level interventions like well-being apps are unlikely to move the needle unless they're paired with changes to how work itself is structured. Real care means becoming what researchers call a behavioral architect: building support into flexibility, manager training, and peer networks, not just into a benefits menu.

But the core of what needs to be fixed is how we define work itself. A recent meta-analysis of 60 years of workplace stress research, spanning nearly 788,000 employees, found that role ambiguity does the most damage to engagement, performance, and trust. Role conflict, incompatible demands and competing priorities, does the most damage to burnout and psychological distress. AI is capable of creating both.

The ambiguity is easy to see. AI is driving more of it every day. Ninety-eight percent of executives are planning organization changes in the next two years, and 99% expect AI to cause at least some headcount reduction, according to Marsh's 2026 Global Talent Trends study. Yet only 32% of executives believe their workforce can combine human and machine capability well. Roles are shifting under people in real time, while confidence remains low at every level about what the changes will actually require.

The conflict is easier to miss. Workers are told AI will make them faster, and are held to that standard. At the same time, they are expected to apply the same level of judgment and quality as before, sometimes even more. But that requires time and is in direct conflict with the "faster" expectation. This contradiction in expectations is what the research says drives people toward burnout and the door.

That contradiction plays out most directly in the work of overseeing AI itself. Many roles are shifting away from human execution and toward AI oversight, and overseeing AI's output is exactly where the demand for speed and the demand for judgment collide. This kind of work redesign carries measurable risk. BCG's 2026 study of 1,488 workers found that workers with high oversight loads report 12% more mental fatigue, and employees who believe AI has quietly raised their workload score 12% higher on fatigue too.

Workers are not only being stressed by ambiguity, conflict, and oversight fatigue but also fear losing their jobs. ADP's 2025 Global Workforce Survey found only 22% of workers worldwide strongly agree their job is safe from elimination. But workers who felt their employer was investing in their skills were 5.3 times more likely to feel secure in their job. That number doesn't move because someone announced a benefit. It moves because someone defined a path forward.

All of this points in one direction: agile job architecture that cuts through the ambiguity and resolves the conflict.

Kevin Cox, who spent his career as CHRO of GE, American Express, and Pepsi Bottling Group, puts this in practitioner terms. He describes the current mandate as optimizing how people and AI work together, and sees CHROs as responsible for deciding who does what work, where it happens, and how. He also carries a line from his former boss at American Express, Ken Chenault, about a leader's job being to define reality and give hope. And he believes part of a CHRO's role is making sure the organization actually lives up to the culture it claims:

“I think of CHROs as the primary architects of a company's operating model.”Kevin Cox, former CHRO of GE, American Express, and Pepsi Bottling Group

This is the reframe: employers are treating mental health concerns by doubling down on traditional methods to support it. These help, but they are only part of the solution.

When AI changes what a role requires and no one rebuilds the architecture to reflect it, people are left to guess if their skills are still relevant. That guessing is the stress, and it's on top of everything else already weighing on them.

The fix has two parts, and both are necessary. First, real investment in mental health support, that is bespoke and meets people where they are. Second, job architecture rebuilt at the pace AI is rebuilding work: clear but flexible job levels, expectations that don't pull people in two directions at once, rewards adapted for how AI changes each role, defined governance so there are guardrails on oversight overload, and a visible path for how skills translate into career growth. Give people care and a map, and the anxiety has somewhere to go besides inward.

Qspired.
We're built for what's next in work.