Getting AI Right in Home Care: Supporting Workers, Not Replacing Them

A groundbreaking three-part report series authored by The CareWorks Project CEO Robert Espinoza and produced by the National Council on Aging, in partnership with the Administration for Community Living, tackles the complicated but critical topic of AI use in home care.

Artificial intelligence is rapidly changing how people work, communicate, and access information, and home care is no exception.

A recent three-part report series from the National Council on Aging’s (NCOA) Direct Care Workforce Strategies Center examines what this transformation could mean for home care workers and the organizations that employ them. The Center is generously funded by the Administration for Community Living (ACL).

Together, the reports provide a foundational overview of AI in home care; examine how agencies and technology innovators are using AI to support 40 common responsibilities of home care workers and their employers, 20 for each; and bring workers and other experts into the conversation about AI’s current use, future potential, and necessary safeguards. The result is a timely contribution to a field confronting two realities at once: enormous workforce challenges and extraordinarily rapid technological change.

The CareWorks Project was honored to partner with NCOA to develop the series. Here, Robert Espinoza, Founder CEO of The CareWorks Project and author of the three reports, speaks with Nicole Howell, Director of Direct Care Workforce Development at NCOA, about her history with these issues, what the research reveals about AI’s potential in home care, and what it will take to ensure these technologies strengthen the people and relationships at the heart of care.

Robert Espinoza: Nicole, before we get into AI and home care, I’d love to start more personally. What first drew you into this work around aging, long-term care, and the direct care workforce? And what continues to keep you committed to it today?

Nicole Howell: I came into this work through aging services and spent many years leading long-term care ombudsman programs. In that role, my job was to advocate for people receiving long-term care, and I continually saw that many of the challenges we were trying to solve were actually workforce challenges.

Whether it was quality of care, consistency, access, or a person’s ability to remain in the setting they chose, there was so often a workforce issue underneath it. There simply were not enough workers, turnover was high, and workers were being asked to do incredibly difficult and important jobs without the pay, training, support, or career opportunities they needed.

One of the first workforce efforts I helped develop was a local training program for CNAs that brought together key community stakeholders. That experience really stayed with me because I saw how much more we could accomplish when employers, educators, advocates, and others came together around a shared problem. It showed me early on that workforce challenges are not going to be solved by any one organization or program.

That has shaped how I approach this work today, including through the Direct Care Workforce Strategies Center. A big part of the Center’s work is bringing together states, workers, providers, national partners, and other stakeholders to solve problems collaboratively and turn ideas into action.

What keeps me committed is that this is a problem we can solve. We know many of the challenges facing the workforce, and across the country states, providers, workers, families, and policymakers are testing new approaches and finding ways to do things differently. There is a real opportunity to build a stronger workforce while also improving the experience of care for older adults, people with disabilities, and their families.

Robert: Let’s turn to the new report series on AI and home care. When we first began discussing this project last fall, I was struck by how little the long-term care sector had collectively grappled with AI, despite the speed at which these technologies are evolving. As I conducted the research and interviews, it became increasingly clear that AI is already entering home care systems, whether the field feels prepared or not.

The reports offered an opportunity to begin a more intentional conversation before these technologies become fully embedded. From NCOA’s perspective, why was it important to begin exploring this issue now?

Nicole: It felt important to explore this now because AI is already becoming part of how organizations operate, how workers do their jobs, and how people access information and services. Home care is not going to be separate from that broader shift.

At the same time, caregiving is already dealing with significant workforce shortages, high turnover, fragmented systems, and growing demand for care. That makes it even more important that we are thoughtful about how these tools are introduced and what problems we are actually asking them to solve.

For NCOA and the Strategies Center, we saw an opportunity to help the field get ahead of the conversation. Rather than waiting until AI is deeply embedded and then reacting to the consequences, we wanted to better understand how it is already being used, where it could genuinely support workers and providers, and where we need stronger safeguards.

Most importantly, we wanted to make sure the people closest to care are part of shaping what comes next. If workers, older adults, people with disabilities, families, and providers are not part of these conversations early, we risk building technology around the system rather than around the people who actually experience it.

Robert: Public discourse around AI often swings between utopian promises and existential panic. That binary didn’t seem especially useful for aging, disability, and long-term care, where AI is entering systems already defined by limited resources, workforce strain, and enormous complexity.

The reports therefore ask a harder question: How do we shape these tools responsibly in a sector already under enormous strain? Why did that framing feel important to NCOA and ACL?

Nicole: That framing felt important because AI is not entering a perfect system. It is entering a sector already facing workforce shortages, high turnover, limited resources, fragmented systems, and growing demand for care.

For NCOA, the question was never simply whether AI is good or bad. The more important question is how these tools are developed and used, and whether they actually make the system work better for workers and the people receiving care.

If we introduce technology without addressing the realities of the workforce, we risk simply layering new tools on top of existing challenges. But if we are intentional, AI has the potential to reduce administrative burden, improve access to information and training, strengthen communication, and give workers more time to focus on care.

I also think this is a real opportunity for our sector to be forward thinking. Too often, aging and long-term care are put in the position of responding to a problem after it has already taken hold. With AI, we have a chance to help shape how these tools develop and how they are used from the beginning. Caregiving is already a huge part of the U.S. economy, touching workers, families, employers, health care systems, and communities. That gives us both a responsibility and an opportunity to make sure innovation in this space reflects the realities of care and the people who make it possible.

Robert: One thing that stood out in the research and interviews was how much home care workers carry every day, and how invisible much of that burden remains. Documentation, fragmented communication systems, scheduling instability, travel burdens, and, most intensely, poverty-level compensation and limited career opportunities all shape the job.

If AI is going to play a meaningful role in this sector, it should materially improve workers’ experiences, not simply extract more productivity from an already strained workforce. Where do you see the greatest opportunity for AI to genuinely improve workers’ day-to-day experiences, beyond organizational efficiency?

Nicole: I think the greatest opportunity is to take some of the friction out of the job. Direct care workers are doing incredibly demanding work, but so much of their day can also be consumed by documentation, scheduling, communication challenges, travel, and trying to navigate systems that do not always work well together.

If AI can reduce some of that administrative burden, help workers get information more quickly, make scheduling more predictable, or improve communication across the care team, that can have a real impact on their day-to-day experience.

I am also really interested in the potential around training and career development. AI could make training more accessible, more personalized, and available when workers actually need it. It could help workers build skills, see clearer career pathways, and access information in different languages or formats.

There is also a real opportunity to think about the next generation of caregivers. Gen Alpha is growing up with AI as an extension of everyday life. If we want to attract younger workers into caregiving careers, we should be thinking now about how technology can make these jobs more connected, more supportive, and more aligned with how the next generation expects to learn and work.

Robert: One of the strongest themes across the series is the distinction between replacing workers and augmenting them. Much of the public imagination around AI centers on replacement, understandably so given the recurring stories about layoffs and automation. But home care is deeply relational work.

What interested me here was whether technology could reduce the friction surrounding care while preserving, or even strengthening, the humanity at its center. Why was it important for the reports to emphasize support, workflow improvement, and burden reduction rather than automation alone?

Nicole: One of the questions I heard a lot when we started this work was, “Does this mean robots are going to be caring for us?” And the answer is much more nuanced than that.

When we looked at where AI could have the greatest value in home care, it was not in replacing the worker. It was in addressing all of the things around the worker that make an already difficult job harder. Sort of, let humans do human things and let technology enable more of that connection.

For me, that is the real opportunity. Use technology to reduce the burden around care so workers have more time and capacity for the parts of the job that only people can do.

Robert: The research also kept bringing me back to the invisibility of home care work. At The CareWorks Project, we think a great deal about how long-term care systems can make workers simultaneously essential and invisible. AI could deepen that dynamic through greater extraction and surveillance, or it could help build systems that better recognize and support workers. Implementation choices matter enormously.

What determines whether workers experience AI as something that makes them feel more supported and valued, versus more surveilled and controlled?

Nicole: I think it starts with the purpose of the technology and who is involved in deciding how it is used.

If AI is introduced primarily to monitor workers, track productivity, or squeeze more out of an already stretched workforce, workers are going to experience that very differently than a tool designed to make their jobs easier and give them better support.

Workers also have to be part of the conversation from the beginning. They know where the real friction is in their jobs, where technology could help, and where it could create new problems. Too often, technology is purchased and implemented first, and workers are asked for feedback later.

And ultimately, I look to the workers to tell me how they feel. Do they feel more supported? Do they have more time to focus on care? Do they feel like the technology is helping them do their jobs better, or do they feel watched and controlled? Those experiences matter.

Transparency matters too. Workers should understand what information is being collected, how it is being used, who has access to it, and whether it could affect scheduling, performance reviews, or employment decisions.

For me, the difference between support and surveillance comes down to whether technology is being used with workers or being done to workers.

Stay tuned for part two: Getting AI Right in Home Care: Putting People, Values, and Safeguards First.