Sensi.ai helps older adults keep living at home, but the price is near-constant listening

Demand is growing in the US for AI devices that help older adults stay in their own homes longer instead of moving into assisted living. The story of Sensi.ai shows why families and care agencies are willing to put a "smart" microphone in someone's home, even if the price is near-constant listening.

How Sensi works

Sensi.ai is a small AI-powered home sensor that sits under a table or next to an armchair. Unlike voice assistants, it doesn't wait for someone to say "help." It listens to ambient sound and tracks signs of risk: a fall, a scream, a heavy cough, unusual activity, frequent trips to the bathroom, or remarks about having trouble walking. When the system detects a dangerous episode, it sends an alert to relatives or care agency staff.

The company claims its algorithm can spot deviations from a person's usual routine and even signs of loneliness or cognitive decline. In the real case behind this story, the device recorded not only noises but also fragments of an elderly man's private conversations. When his family requested the transcripts, they learned the system forwarded them to caregivers whenever it heard him mention falling or being unsteady on his feet.

Why the market is growing

The idea is appealing because families and care agencies have few good alternatives. For many people, moving into a nursing home is too expensive and psychologically unacceptable, while the home care market is strained by staff shortages. Against this backdrop, these devices are sold not as just another gadget but as a compromise between independence and safety.

• Sensi.ai has raised $100 million in funding and says it works with 80% of the largest home care chains in North America.
• The company promises around 90% accuracy, though staff who use it admit the system can mistake a person falling for a dropped remote.
• A private room in a US nursing home costs more than $108,000 a year on average, so families are looking for ways to keep relatives at home longer.
• By one estimate, the US care system will need to fill more than 9 million jobs by 2031.
• A quarter of people caring for loved ones already use remote monitoring through apps, wearables and other systems.

For agencies, it's also a business tool: the device helps them respond to incidents faster, retain clients and expand the number of billable care hours. The reporting cites cases where the system genuinely helped: it detected a fall near a toilet and picked up a cough early enough for a staff member to check on the client.

What worries doctors

But the list of questions is growing along with the benefits. Neurologists who use AI to diagnose dementia are skeptical of claims that cognitive decline can be detected from speech and everyday behavior. Sensi.ai has not yet received FDA clearance for such claims, and it discloses almost nothing publicly about what data its models were trained on or how it checks for false positives.

Consent is an equally pressing issue. Older adults often agree to the sensor under pressure from family, or simply forget what they signed in the privacy form. Care ethics researchers warn that the home is gradually becoming a space where surveillance is a condition of receiving help rather than a free choice.

"Consent by itself doesn't make the process ethical if a person is offered two bad options," says researcher

Clara Berridge.

There's also a more personal layer to the problem. These systems promise loved ones peace of mind from a distance, but they also hand over access to deeply intimate data: conversations, loneliness, fears, frailty. What looks like a "risk signal" to the platform may be, for the person themselves, the last bit of privacy left in their own home.

What it means

AI-based monitoring of older adults is quickly moving out of its niche and becoming part of the core infrastructure of care. Demand will only grow: families want to keep relatives at home, agencies are short on staff, and government regulation isn't keeping pace with the market. The key question is no longer whether these devices will be everywhere, but who will be listening to a person's life inside their own home, and under what rules.