Skip to content

Artificial Intelligence (AI) skilled companion for people in assisted living facilities: Pilot study

The effect of an AI skilled companion on loneliness for people in assisted living facilities: Pilot study

Status
Not yet recruiting
Phases
Unknown
Study type
Interventional
Source
ANZCTR
Registry ID
ACTRN12625000500459
Enrollment
15
Registered
2025-05-22
Start date
2025-07-01
Completion date
Unknown
Last updated
2025-09-08

For informational purposes only — not medical advice. Sourced from public registries and may not reflect the latest updates. Terms

Conditions

None listed

Brief summary

Older adults and people living with dementia in assisted living facilities often feel lonely. In 2017 (pre- covid) it was suggested that 40% of residents seldom or never receive visits from family or friends. Staff workloads are frequently high leaving staff little time to chat to residents. Residents may not have anyone to speak to or have conversations with for long periods of time. In one study less than a third of residents had a friend in the care facility and residents living with mild cognitive impairment or dementia may find it difficult to sustain relationships or engage meaningfully with other residents. This may lead to more feelings of loneliness with associated impact being depression, anxiety, emotional distress and high blood pressure, obesity, heart disease. We have developed an embodied, screen-based, artificial intelligence (AI) and lived experience supported, digital companion to interact with people and support naturalistic conversations. The digital companion for this study has been developed using lived experience data from older adults, and people living with MCI and dementia. We will carry out a pilot study in an assisted living facility making the digital companion available to residents on large iPads on adjustable movable stands positioned at head height (so the iPads do not have to be held and allow for natural face to face interactions) and also a large human-sized screen in a common area. Based on iterative testing throughout development and existing user testing, we expect to see that people find it easy to speak to the AI character, and that these conversations reduce feelings of loneliness, depressive symptoms and changed behaviours (such as agitation) over time.

Interventions

Participants will interact with an agentic AI companion. They will be provided supervised access at least twice a week for 15 minutes during the 4 week intervention. They will also have unsupervised access available during the intervention on-demand. The participants will access a purpose-built AI companion access on a touch-screen tablet and also on a large human-sized screen in a common area. The participants will have conversations with the AI companion, which will reply based on its programm

Participants will interact with an agentic AI companion. They will be provided supervised access at least twice a week for 15 minutes during the 4 week intervention. They will also have unsupervised access available during the intervention on-demand. The participants will access a purpose-built AI companion access on a touch-screen tablet and also on a large human-sized screen in a common area. The participants will have conversations with the AI companion, which will reply based on its programming in a naturalistic manner (not pre-scripted responses). The AI companion has been previously co-designed with people with dementia and is called 'Viv'. There is no maximum number of times the participant can access the companion per week.

Sponsors

UNSW Sydney
Lead SponsorUniversity

Study design

Allocation
Non-randomised trial
Intervention model
Single group
Primary purpose
Treatment
Masking
Open (masking not used)

Eligibility

Sex/Gender
All
Age
55 Years to No maximum
Healthy volunteers
Yes

Inclusion criteria

People in assisted living facilities: Aged 55 or older Living in an assisted living facility Cognitively healthy, or living with mild cognitive impairment, or mild to moderate dementia Study partner Inclusion criteria include: Staff member at an assisted living facility or care partner supporting a person participating in the study.

Exclusion criteria

People in assisted living facilities: Not able to give informed consent or have family member or guardian provide consent Study partner: N/A

Outcome results

None listed

Source: ANZCTR · Data processed: Feb 4, 2026