The Skill That Makes Women AI-Proof Is the Same Skill That’s Burning Them Out

Women at work spend hours a week listening, reassuring, and steadying their colleagues who are coming apart. Those same women pay a price.

The Skill That Makes Women AI-Proof Is the Same Skill That’s Burning Them Out
Artwork by Hanna Melin.
The Persistent is available as a newsletter. Sign up here to get it delivered to your inbox. 💛

“Amy” is the head of marketing at a small company. She has one direct report, and we’ll call him “Sam.” Sam is a nervous wreck. As executives at their firm scramble to get the firm "AI-ready," they keep changing their minds about what that actually means. Use AI for this, not that. Get up to speed, but don’t go too fast. Somewhere in the noise, Sam has decided he'll be out of a job by fall, and he carries that dread to Amy's virtual desk several times a week, looking for reassurance. 

The problem is that Amy, like every manager across America, has no idea what to tell him. She helps soothe Sam’s anxieties anyway.

Perhaps you know a “Sam.” Perhaps you are an “Amy.”

While Amy and Sam's names aren't real, they are two very real office personality types living out a scene that is unfolding across countless companies right now. 

The truth is, workers like “Sam” have every reason to be rattled. A recent workplace report from Modern Health found that 69% of U.S. employees expect AI to trigger layoffs at their company within three years. And only 22% of workers in ADP's latest global survey of nearly 40,000 people strongly agreed that their own job was safe from elimination more broadly.  

The problem is that Sam’s worries have to go somewhere, and more often than not, there’s an “Amy” in the workplace — almost always a woman — who’s absorbing all that angst.

Managers like Amy, in our hypothetical above, provide calm reassurance the way women have always done — imagine a mom smoothing her sick child’s brow — while the men alongside them at work provide comparatively little of it. 

But Sam’s also not the only one with fear. No doubt Amy is also feeling angsty, but only Sam gets his fear assuaged. As for Amy, she just has to take care of herself. 

AI Skews Beautiful. What Is That Doing to Us?
Even when creators want them to be ugly, images of people generated by AI tend to be gorgeous,

It’s called the ‘care tax’

It turns out this lopsided arrangement already has a name. It’s what Harvard Business School researchers Colleen Ammerman and Deepa Purushothaman call the “care tax,” and senior-level women are footing the biggest bill. In their survey of more than 350 women in management, 82%  said they spend at least 30% of their week on caring work: listening, reassuring, steadying whoever is coming apart. And much of it goes unseen by the people not doing it. Ammerman and Purushothaman found that men often didn't register the labor happening around them at all, while 76% of women said this work in their organization falls mostly to women.

When I asked Purushothaman what's behind the growing load, she pointed straight at AI. The shift is recent, she said. When she and Ammerman started running the survey last fall, women tied their uncertainty to tariffs and the economy, but as she continued to gather comments on LinkedIn and elsewhere, AI has become front and center. As one respondent put it, "AI is the biggest source of uncertainty in my workplace."

Employees bring their fear about the technology to the women around them, and those women absorb it. "Women are becoming the modern infrastructure," Purushothaman told me. What she means, in plain terms: women have become the people who keep everyone else at work functioning — the ones colleagues lean on to stay calm and keep going — during one of the biggest tech upheavals the workplace has ever faced. And nearly 59% of the women surveyed said the caring load keeps growing.

Why Women Don’t Trust AI
It’s less about safety, more about the ethical consequences.

All compliment, no comp

Plenty has been written about women and the AI divide, and the headlines keep coming: The adoption gap is closing: women now use AI chatbots at nearly the same rate as men, but the shaming hasn't let up. Women get scolded for lagging on AI and branded "cheats" when they lean in, a tidy way to make sure they lose both ways. And when women do hold back, the reason is less about skill or access than about how clearly they read the technology's risks.

But the counter-narrative to all of this is rosier, and it goes like this: Women will win the AI era precisely because of what they bring to the table that machines can't. In a word, it’s empathy. 

Research this year from LinkedIn found that as AI literacy turns into table stakes, employers are placing even greater value on human capabilities like empathy and personal connection. Michael Platt, a neuroscientist at the Wharton School, calls these the capacities AI "may never master” while Lean In's latest newsletter makes the case that women's empathetic leadership style is their real edge in an AI-driven economy. 

If that sounds like good news, surprise, there’s a catch. The skills these experts are betting will carry women through the AI revolution are the same ones their employers won’t pay for. Empathy is the asset everyone celebrates but no one budgets for.

Ammerman is clear that this work isn’t actually hard to value, given the extent to which companies track retention, attrition and engagement all day long.

“It’s not invisible because it’s difficult to identify and price,” she told me. “It’s invisible because it’s women who are doing it.”

Calling empathy a superpower, Ammerman warned, is “one of these compliments that actually just helps lock in these inequalities.” Purushothaman named the mechanism even more plainly: “The more you do it, the more you’re asked to do it.” Get good at holding everyone together, and your reward is a longer line outside your door.

Meanwhile, the broader workplace is already shifting: The LinkedIn research I referenced above projects that women’s jobs will be disrupted by generative AI at higher rates than men’s. In other words, just as women are told that their humanity — their unpaid work — is their hedge against the machines, it’s their paid work that is most exposed to automation, and the most likely to be disrupted or to disappear altogether.

The advantage and the tax turn out to be the same skill. Only one of them shows up on a paycheck. 

Is this all there is?

Ammerman and Purushothaman’s survey results included comments that read less like burnout and more like a reckoning. One respondent wrote that when emotional labor is invisible and unrewarded, but wholly expected of women leaders, it's “easy to see why many highly capable women are having a crisis of meaning. The questions I hear over and over are: Is this all there is? And is it really worth it?"  

Indeed, more than one-third of the women in the study said the caring work makes them more likely to leave their current role. The comments suggest the driver is as much about meaning as exhaustion. 

Too often, women at work are flattered into staying exactly where they are. But we all know flattery doesn’t pay the bills. And that’s how “your empathy is a superpower” lands like a promotion, but works like a leash. 

The women walking out have made their peace with the labor of holding people together. Good for them.

Empathy may well be the skill that outlasts the machines. The question is how women — and anyone else able to nurture it — will be able to profit from it. And how all that talent, now on the way out the door, can be encouraged to return.

When AI is Listening, Women Speak More
A new study shows that when meetings are transcribed by AI, women find it easier to speak.
Caroline Fairchild is a journalist, researcher and public speaker focused on women, power and the future of work. 💛 Born in Sweden, Hanna Melin lives in East London, where she works as an illustrator and a designer. Melin’s work appears in newspapers, magazines, product advertisements, and the commercial sector. She loves drawing people and making ceramic clay figures of cats and dogs.