Intuition Robotic’s ElliQ gained national attention in The New York Times and on its podcast The Daily earlier this year—an indication that ElliQ and artificial intelligence (AI) robotic companions like it are emerging from pilot projects and test uses in limited settings into the broader public view. With it comes a question many LeadingAge members and others are already asking: Should we be paying closer attention to AI companions?

The answer is yes, but with the same rigor and care providers bring to any technology that operates in the daily lives, relationships, and well-being of older adults.

Which Gaps AI Companions Fill

An AI companion is not an app in the traditional sense. It is designed to do something apps rarely do: show up consistently, over time, in a way that feels conversational and relational. Some AI companions are social robots. Some use voice-activated speakers. Some live on a phone or tablet. Some call an older adult directly on the telephone. What they share is a design intent: to check in, respond, encourage, prompt, and engage as part of a daily rhythm.

In practice, that might mean reminding someone about an activity they mentioned wanting to try, asking how a doctor’s appointment went, suggesting a song tied to a memory, encouraging an afternoon walk, or simply being present during the quiet hours that stretch between scheduled programming. For some older adults, those quiet hours make up much of the day.

What the Evidence Is Beginning to Show

RiverSpring Living, a LeadingAge CAST Patron in Riverdale, NY, piloted Meela, a phone-based AI companion designed to call and engage older adults as part of an effort to address social isolation. Staff observed that residents used Meela in deeply human ways: talking about baseball, reading plays together, and reminiscing. The technology faded into the background. The connection was what mattered.

RiverSpring also collaborated on a feasibility study, published in the Journal of the American Medical Directors Association, examining an AI-driven voice companion in post-acute and long-term care settings. The preliminary findings were encouraging. Participants showed reductions in depressive and anxiety symptoms, with larger improvements among those who had higher symptom levels at the start of the study. For participants with more significant depressive symptoms, scores declined by an average of 5.7 points on the Patient Health Questionnaire-9 (PHQ-9), while anxiety scores declined overall on the Generalized Anxiety Disorder-7 (GAD-7) measure. The authors appropriately cautioned that larger, controlled studies are still needed, but the results point to the potential for AI companions to support emotional well-being when used thoughtfully.

Why This Category Deserves Serious Attention

AI companions do not solve loneliness, and no responsible provider should frame them that way. The case for AI companions is more specific. Connection is unevenly distributed in the lives of older adults, and that unevenness is often invisible from the outside.

Some people have family nearby, active social networks, regular appointments and classes, transportation, and daily touchpoints with other residents, staff, and neighbors. Others have far fewer interactions built into their day. Providers know who those people are. They also know that even the strongest programming cannot fill every gap for every person, every day.

That gap is where AI companions may have a genuine role. While the value will not be the same for every user, in general they add an extra layer of support, particularly for older adults who benefit from reinforced routines, encouragement to participate, or support for wellness goals. For care teams, these tools may also surface patterns in engagement, mood, or daily rhythm that are difficult to observe at scale.

Where Caution Is Warranted

The quality that makes AI companions potentially valuable—the ability to build familiarity and trust over time—is also what demands the most careful oversight.

A tool that sounds warm and friendly can build trust quickly, often in a space that is deeply personal. Older adults deserve to know, clearly and consistently, when they are interacting with AI. Families and staff need to understand what a given tool does, what it cannot do, what data it collects, and how that information is used. Providers will need to think carefully about consent, cognitive status, user expectations, and whether a particular tool is genuinely strengthening connection or simply adding another device to someone’s life.

Privacy and data practices deserve particular scrutiny. Conversations between a person and an AI companion may include health information, emotional disclosures, and personal history. Providers should ask hard questions before any implementation about where that data goes, who has access, and how it is protected.

A Framework for Getting Started

For most aging services organizations, the right next step is not rapid adoption. It is thoughtful evaluation. That means starting with a small group of older adults most likely to benefit, listening closely to their experience, involving direct care staff and family members where appropriate, and asking honestly whether the tool is helping someone feel more connected, more engaged, or more supported. It also means being willing to stop if the answer is no.

As the vendor landscape continues to shift, providers should focus less on any one product and more on whether the approach improves connection, engagement, or support for older adults. To help providers sort through this ever-changing technology landscape in order to better support residents, CAST regularly updates its Social Connectedness and Engagement Techology Selection Tool, which includes information on ElliQ and other products.

The Bigger Picture

AI companions deserve a place in the conversation aging services is having about technology, connection, and scale. They are not a shortcut, and they are not the solution to loneliness, social isolation, or workforce challenges. But for specific individuals, in specific circumstances, they may be a meaningful addition to a broader ecosystem of connection and support.

That ecosystem still needs people at its center: staff who know residents and clients as individuals, families who remain part of daily life, volunteers and community partners who extend reach, programs that give people purpose, and technology that helps fill gaps without pretending to replace the relationships people need most.

Used thoughtfully, AI companions can help aging services providers reach older adults more consistently, notice needs earlier, and add another layer of connection. The bigger opportunity is not to replace human relationships with artificial ones, but to use every available tool to help more older adults feel known, supported, and included.