A new study by Lucia Vitale, a doctoral candidate in the Department of Political Science at the University of California, Santa Cruz, examines the current mix of promise and uncertainty as artificial intelligence-powered tools increasingly permeate healthcare settings. There is.
Proponents of AI say the technology can help manage medical supply chains, monitor disease outbreaks, diagnose and interpret medical images, and even address equity disparities in access to health care by filling gaps in health care workers. is intended to reduce. However, there are concerns about privacy rights, racial and gender bias in models, lack of transparency in AI decision-making processes that can lead to errors in patient care, and even the potential for insurance companies to use AI to discriminate against people. Some people are sounding the alarm on issues such as sexuality. Poor health.
The type of impact these tools ultimately have depends on how they are developed and deployed. In a paper published in Social Science & Medicine, Vitale and co-author Leah Shipton, a doctoral candidate at the University of British Columbia, provide an extensive literature analysis of the current trajectory of AI in healthcare. I did it. They argue that AI will ultimately have only limited impact because it engages in a “politics of avoidance” that distracts from and even exacerbates more fundamental structural problems in global public health. It claims to be in a position to be the latest in a long line of technological advances. .
For example, like many previous technological interventions, most AI being developed for health focuses on treating disease, ignoring the fundamental determinants of health. Vitale and Shipton say the hype around unproven AI tools has distracted from the urgent need to implement low-tech, evidence-based, holistic interventions, such as community health workers and harm reduction programs. I’m worried that it will deviate.
I’ve seen this pattern before. We continue to invest in these technological silver bullets, but they fail to actually change public health. This is because these policies do not address the deep political and social determinants of health, ranging from health policy priorities to access to healthy food and safe spaces. To live. ”
Lucia Vitale, Department of Political Science, University of California, Santa Cruz
AI also has the potential to continue or exacerbate patterns of harm and exploitation that have been historically common in the biopharmaceutical industry. One example discussed in the paper is that ownership of and profits from AI are currently concentrated in high-income countries, while low- and middle-income countries with weaker regulation are less likely to be able to extract data or potentially risk There is a possibility that it could become the subject of experiments to introduce new artificial intelligence with high performance. technology.
The paper also predicts that weak regulation of AI will continue to prioritize intellectual property rights and industry incentives over fair and affordable public access to new treatments and tools. And as corporate profit motives continue to drive product development, AI companies will also continue to ignore the needs of the world’s poorest people when deciding which problems to invest in R&D. likely to follow long-term trends in the technology sector.
But Vitale and Shipton identified a bright spot. AI has the potential to disrupt conventional wisdom and create deeper impact by focusing on improving the healthcare system itself. AI can be used to allocate resources more efficiently and triage patients more effectively across hospitals. Diagnostic tools have the potential to improve efficiency and expand the capacity of general practitioners in small, rural hospitals without specialists. AI may even be able to provide basic but essential medical services to bridge workforce and specialty gaps, such as providing prenatal diagnostics in areas with growing obstetric deserts. .
All of these applications have the potential to result in more equitable access to care. However, the results are not guaranteed. Depending on how and where these technologies are deployed, they may succeed in filling gaps in healthcare where there is a true shortage of healthcare workers, or they may lead to unemployment and precarious casual work for existing healthcare workers. may lead to. And unless the root causes of health worker shortages, such as burnout and “brain drain” to high-income countries, are addressed, communities will still lack the diagnostic and spread-detection capacity to use AI. Tools may end up providing useless diagnostics and outbreak detection. respond.
To maximize benefits and minimize harm, Vitale and Shipton argue that regulation needs to be put in place before AI expands further into the medical field. With the right safeguards, AI could avoid following the harmful patterns of the past and instead chart a new path that ensures future projects align with the public interest.
“Thanks to AI, we have the opportunity to modify the way we manage new technologies,” Shipton said. “However, we need a clear agenda and framework for the ethical governance of AI medical technologies through the World Health Organization, major public-private partnerships that fund and deliver medical interventions, and countries such as the US, India, and China that have technology companies.” Doing so will require continued civil society advocacy. ”
sauce:
University of California Santa Cruz
Reference magazines:
Shipton, L., Vitale, L. (2024) Artificial intelligence and the politics of avoidance in global health. Social science and medicine. doi.org/10.1016/j.socscimed.2024.117274.
