In late February, Fortune ran a headline that captured a shift in national mood: “The week the AI scare turned real and America realized maybe it isn’t ready for what’s coming.” Tech companies had been citing AI-driven efficiencies during earnings calls for months. Now the layoffs were accelerating. Amazon alone eliminated roughly 16,000 corporate positions in the first quarter of 2026, flattening management layers to fund AI infrastructure. The numbers are startling.

But they describe one thing: jobs lost. For families, the losses multiply.

The first domino

Layoffs are about more than jobs

Sheila, a project manager in Charlotte, North Carolina, sees a new meeting on her calendar on a Friday afternoon. Her stomach sinks. She knows what’s coming: her company has been automating many positions, and her role is going to be eliminated. 

Sheila’s job is one data point in a layoff tracker. But for her, that Friday meeting is the first domino. Her family’s health care coverage ran through her employer, so suddenly she’s looking at thousands of dollars per month in COBRA payments. Her partner starts scrambling for benefits-eligible work. Child care arrangements built around two incomes fall apart. Sheila’s parents, who depend on her for financial help, start making hard decisions about prescriptions and bills.

One job eliminated. Six lives across three generations changed.

“Job displacement” implies a clean swap where a role disappears, then a worker moves to a new one.

What’s happening here—and to thousands of families like Sheila’s—falls in the space between workforce policy and family policy, where neither field claims responsibility. I’m calling it a compound displacement, because a single job loss never stays single. That’s what I’ll be investigating over the next year as Capita’s Future of Work Senior Fellow.

“Job displacement” implies a clean swap where a role disappears, then a worker moves to a new one. Compound displacement takes a broader view. It asks, Can this family hold together through the transition? That depends on structures and relationships that don’t show up on a resume…or in the worker surveys that most workforce policy is built on. It depends on how the income, benefits, and caregiving are distributed across the household before the disruption hits. 

Economic resilience belongs to the family, not the individual. 

Across the same round of layoffs that sent Sheila’s family reeling, some families are able to absorb the hit. There may be a partner who also carries benefits. Or grandparents close enough to step in on child care. Perhaps there are three months of savings, rather than three weeks, to better cushion the blow. Same displacement. Different outcome. The difference has little to do with the initial job loss or the person who lost the job.

What separates the outcomes for these families—crisis or recovery—was how income, benefits, and caregiving were distributed across the family before the disruption hit. Economic resilience belongs to the family, not the individual.

AI-driven layoffs affect women the hardest, with impacts across households

Over the last half-century, economic risk has steadily shifted away from institutions—employers, unions, the welfare state—and onto families. AI disruption is accelerating that transfer. And the families absorbing the most risk are the ones least equipped to carry it.

AI exposure doesn’t guarantee displacement. But it shows where automation pressure is landing first, and who has the least room to absorb it.

A 2026 ILO report found that women’s occupations are three times more likely to fall into the highest AI exposure categories. And a recent Anthropic study measuring actual AI use—not forecasts, but how AI is already being deployed—across 800 occupations confirms the pattern: the most AI-exposed jobs are disproportionately held by women. Data from the National Bureau of Economic Research sharpens the picture: 86% of workers with high AI exposure and low capacity to adapt are women. Health care administration. Education support. Human resources. Marketing. Project management.  

For many women (especially primary earners) that job is the load-bearing wall of the household.

These are not marginal jobs. They are middle-income, credentialed, professional roles. The kind that you’d think are “safe.” AI exposure doesn’t guarantee displacement. But it shows where automation pressure is landing first, and who has the least room to absorb it.

For many women (especially primary earners) that job is the load-bearing wall of the household. It carries the benefits. It anchors the budget. It keeps everything standing. When it goes, the whole structure shifts.

What’s more urgent is that the stakes are not evenly distributed. Median wealth for Black families stands at $44,900, compared to $285,000 for white families. More than 80% of Black mothers are breadwinners: sole, primary, or co-earners holding their families’ health care, income, and child care—their economic lives—together. There’s no margin.

So much of what we read about AI and work focuses on the individual. Will workers reskill? Are they adaptable? Do they have a growth mindset? These questions skip over the structural reality that the workers most exposed to AI disruption are disproportionately women. And that women are disproportionately the ones managing the household conditions that make recovery possible or impossible. 

The policy reflex is simple: reskill. But reskilling into what, and through which programs? For most workers, the infrastructure doesn’t exist yet. And even where it does, reskilling takes time, money, and bandwidth. Someone has to cover child care. Health insurance has to come from somewhere. The mortgage doesn’t pause. Whether a worker can access a reskilling program (and finish it) depends on household conditions that policy almost never measures.

I want to be precise. Individual adaptability matters. Workers right now are reskilling and landing in stronger positions than before. That’s real. But the current landscape for most AI-impacted workers is an array of certifications, too many browser tabs open, and a vague gesture in the direction of LinkedIn Learning. Navigating this landscape, staying in a program, and surviving the income gap—that depends on household conditions no reskilling platform can address. We’ve been measuring the worker. We haven’t been measuring the ground they’re standing on.

Sheila knows this. Three months after the layoff, she finds a reskilling program that could lead somewhere. But it’s 12 weeks, full time, and unpaid. Her partner just picked up a second shift to cover the insurance gap. Who will be home when her daughters get off the bus? The program has a 40% completion rate. She doesn’t need to ask why.

The Future of Work Fellowship

I’ve spent the last three years working in career navigation, holding hundreds of conversations with people figuring out their next move. The conversations always begin with the individual: their skills, their prospects, the role they should target next. And then, almost without exception, the household enters the frame. The partner’s job security. The child care math. How long they can afford to be in transition. Career decisions aren’t individual optimization problems. They’re household negotiations.

I know this calculus. I have two young children and I’ve made every one of these calculations myself. This isn’t an abstraction to me. It’s Tuesday. 

That’s the central insight behind this fellowship: economic resilience lives at the household level.

The goal of this fellowship isn’t just to document what’s happening. It’s to build the Career Continuity Framework: a practical tool that maps the household-level factors that determine whether a displaced worker can actually recover. If economic resilience is a property of the household, not the individual, we need to measure it that way. 

Over the next 12 months, I’ll conduct 50 to 75 in-depth interviews with working parents in North Carolina navigating AI disruption in middle-income roles. Not surveys. Conversations. I want to hear what’s actually happening before I match it to the data. For a subset of families, I’ll interview a second household member. Resilience is relational, so you can’t measure it by talking to one person. I’ll be grounding the qualitative findings in data on employment transitions and analysis of automation exposure across the state.

The research will focus on North Carolina because the state is a microcosm of national workforce transformation: Research Triangle tech, Charlotte financial services, Piedmont manufacturing—each faces AI disruption differently. Governor Josh Stein established an AI Leadership Council in September 2026, signaling that the state sees what’s coming. If this framework succeeds, North Carolina could be the first state to pilot workforce policy that measures household resilience instead of just individual employability. And I grew up in Charlotte. My family is there. That gives me access and a reason to get this right.

But we already know enough to name what should change. Reskilling programs should ask about access to child care before workers enroll, not after they drop out. Unemployment benefits should account for household structure, not just individuals’ earnings history. Workforce policy should measure what actually predicts whether a family survives a transition. And employers funding AI infrastructure by eliminating roles have a question to answer too: If you’re transferring economic risk to households, what’s your obligation to the ecosystems those workers hold together?

Reskilling programs should ask about access to child care before workers enroll, not after they drop out. Unemployment benefits should account for household structure, not just individuals’ earnings history.

Workforce policy lives in one silo. Family policy lives in another. But what happens to Sheila—the months when a layoff becomes long-term unemployment, when a grandparent moves in to cover child care, when a family restructures itself around a gap that no program exists to fill—falls in the space between. AI is rewriting the economics of both fields simultaneously. This fellowship sits in that gap, because that’s where families are.

Capita’s Family Policy Lab is asking what it would take for American family policy to meet this moment. 

Logan Currie is a Future of Work Senior Fellow at Capita, wheare she studies how AI-driven job displacement disproportionately impacts women and their families. She co-founded Careerspan, a career navigation platform, and has spent three years in conversation with workers figuring out what comes next. She writes about AI and work at Career Security Field Notes on Substack. She holds an MEd from Harvard and a BA from the University of Pennsylvania.