A peer reviewed interview study of 13 caregivers and healthcare professionals in Switzerland and Germany has found broad interest in GRACE, a voice assistant designed to deliver cognitive stimulation exercises and daily support to people with early-stage dementia, but also identified practical, technical and ethical obstacles to using it inside real care settings. The study, published in Digital Health and led by researchers at the University of St. Gallen and the University of Zurich, is the first to gather structured feedback on GRACE from the people who would actually deploy it, rather than from healthy volunteers testing the device in a laboratory.
Researchers conducted 13 semi-structured interviews between October 2024 and May 2025, with four informal caregivers and nine healthcare professionals. Participants watched a recorded demonstration of the GRACE prototype, which walks a user through an onboarding introduction, a guessing game matching animal sounds to animals, and a short-term memory exercise, before answering questions about its usefulness, usability and fit with existing care routines. Respondents rated the onboarding sequence and memory exercise as the most valuable parts of the interaction, saying the tasks mirrored everyday cognitive demands that people with dementia and their families already face.
The same interviews surfaced resistance to folding GRACE's output directly into clinical or caregiving workflows. Participants raised practical concerns about how the device would be installed and maintained in a person's home, technical concerns about reliability and response delays, and ethical concerns centered on data protection and the risk that an always-on assistant could substitute for human contact and deepen a person's isolation rather than relieve it. The authors write that personalization, ease of use and alignment with existing care processes will determine whether the tool is adopted beyond a research setting.
The findings sit alongside an earlier laboratory usability study of the same assistant, published in JMIR Aging in December 2025, which tested GRACE with healthy adults and reported favorable technical feasibility and acceptance in short, 10 to 15 minute sessions. The new interview study was designed specifically to move past that healthy-volunteer setting and ask the caregivers and clinicians who would be responsible for the device in practice what they thought of it, and their answer was qualified: valuable as a complementary activity, not yet ready to be wired into care decisions.