Barbara Barbosa Neves, Alan Petersen, Mor Vered, Adrian Carter, and Maho Omori
Monash University
‘The aged care system is well behind other sectors in the use and application of technology’ noted the Royal Commission into Aged Care Quality and Safety in 2021. [1 p77] Yet, the Commission also warned that new technological systems and initiatives must ‘identify older people’s needs and preferences.’ [1 p147]
Since the onset of the COVID-19 pandemic, we have seen a renewed drive to introduce emerging technologies, such as Artificial Intelligence (AI), into residential aged care. [2] This drive is fuelled by the promise that AI can help solve the systemic issues affecting the sector, being the ‘future of elder care’. [3] For example, robots and smart voice assistants can help stave off loneliness and social isolation; smart medical systems can improve clinical diagnoses; smart sensors can predict falls and health decline. And while AI can assist with some of these matters, we must examine its limitations and potential harms.
Our new study, published in the Journal of Applied Gerontology, [4] aimed to understand how AI for aged care is imagined, designed, and implemented. For this, we conducted interviews with a range of stakeholders: AI developers, aged care staff, and aged care advocates in Australia. We explored their views about AI for later life, whether as developers of technologies for the sector or as those who have the power to implement or support their use in aged care settings.
We found that, despite good intentions, developers made many promises about what AI could fix in aged care, from loneliness to staff shortages. But these promises were often made without considering the complexity of such environments and its residents. For instance, developers described stereotypical ideas about older residents. They were seen as a homogenous group and as technologically incompetent or disinterested. These ageist ideas illustrate how easily age-related bias can make its way into how we create AI technologies. But aged care staff and advocates held similar views about the technological incapacity or interest of older people. AI can be ageist by design but also by implementation.
Our study shows that, alongside other prejudices pervading AI technologies like gender and racial biases, age is a serious but overlooked factor in research and policy. Understanding how ageism is embedded in both AI systems and in aged care settings is essential to combat it. This requires more institutional and public efforts to challenge simplistic narratives about older people. These narratives can lead to the development and application of technologies that exacerbate existing stereotypes and inequities, as they neglect the diversity of later life and the various needs and aspirations of older people. Indeed, their autonomy and dignity were rarely discussed across the stakeholders in our research. This was clear even when we were talking about AI for surveilling their everyday movements. Older residents were sometimes reduced to sick bodies to be constantly monitored.
As stressed by the Royal Commission, ‘Ageism is a systemic problem in the Australian community that must be addressed’. [1 p75] We cannot forget to address it at all levels, including in relation to how we develop and use technologies for care.
*The views and opinions expressed in Knowledge Blogs are those of the authors and do not necessarily reflect those of ARIIA, Flinders University and/or the Australian Government Department of Health and Aged Care.