Fostering Ethical Adoption of Artificial Intelligence-Enabled Assistive Robots (AIAR) Grounded in Person-Centered Dementia Care

Project Summary/Abstract Dementia poses a significant challenge in the US, with one in three older adults living with the condition. This reality not only incurs staggering healthcare costs, exceeding $345 billion annually, but also imposes considerable socio-economic and psycho-social burdens on People Living with Dementia (PLWD)

and their 11 million unpaid care partners. In this landscape, AI-assisted robot technology (AIAR) emerges as a promising solution, offering to enhance care and improve the quality of life for PLWD and their care partners by enabling them to age safely at home. However, the integration of AIAR into dementia care requires navigating

a complex ethical landscape, especially concerning the preservation of dignity (concerns around dehumanization) and autonomy (privacy and consent), which interact in the engagement of PLWD in the continuous decision-making of AIAR assisted care. Current AIAR implementation often marginalizes PLWD in

decision-making processes due to societal perceptions and stigma, which imply that PLWD lacks the capacity for meaningful decision-making. A care partner's intimate understanding of the PLWD's history, preferences, and personality is vital for interpreting subtle cues and enhancing communication, especially as cognitive

abilities fluctuate. However, this understanding can also impede autonomy if care partners presume to know the PLWD's wishes without involving them in the decision-making process, showing the significance of supporting care partners to learn to engage PLWD in making decisions. Our study aims to address this issue by aligning AIAR technology with person-centered dementia care (PCC)

through a two-stage approach. Aim 1 (Stage 0): Describe Experiences of PLWD, Care partners, and AIAR Developers with AIAR Implementation in Everyday Care Context. Adopting in-depth interviews and non- participant observations guided by our conceptual framework, we will describe the experiences, related ethical

complexities, and decision-making processes associated with AIAR usage. We will also identify features of the AIAR that both impede and facilitate collaborative decision making. Aim 2 (Stage 1a): Co-Design a Robust Intervention Protocol with Preliminary Feasibility and Acceptability. We will engage key stakeholders and expert

panels to co-design an ethically informed intervention model that demonstrates preliminary feasibility and acceptability, setting the stage for subsequent efficacy testing and a pragmatic clinical trial involving PLWD-care partner dyads and AIAR systems. The intervention model aims to prioritize the dignity and autonomy of PLWD

and enhance collaborative decision-making. The study holds the potential to direct dementia care towards a future where technology is utilized with respect, integrity, and a profound understanding of human dignity and autonomy, thereby ensuring the ethical and effective integration of AIAR in dementia care.