102. Navigating the Complexities of Trustworthy AI in Healthcare: Ethical, Legal, Technical, Epistemic, and Societal Aspects
Jean-Christophe Belisle-Pipon, Simon Fraser University; Vardit Ravitsky, Université de Montréal; Yael Bensoussan, University of South Florida;
Burgeoning advancements in artificial intelligence (AI) hold tremendous promise for revolutionizing healthcare in diverse aspects, including diagnosis, treatment, research, and administration. Nevertheless, the design, assessment, and implementation of these technologies introduce multifarious challenges that demand a comprehensive approach to ensure ethically-sound and trustworthy applications in healthcare settings. This open panel solicits papers that encourage interdisciplinary discourse on the legal, ethical, technical, epistemic, and societal dimensions implicated in developing AI technologies that uphold trustworthiness, patient confidentiality, human dignity, and health equity, while circumventing the intensification of pre-existing disparities.
Areas of interest encompass the evidence paradigm generated by AI; reliable causal mechanistic insights; implicit discrimination and prejudiced algorithms; diversity of interests in the development of AI health solutions and the technological imperative; unforeseen issues impacting minority populations; transparency/interpretability/explainability/contestability; trustworthy stewardship and current responsibility gaps. The panel endeavors to highlight the significance of devising governance frameworks and structures, such as compliance and oversight safeguards, which facilitate transparent and equitable access to AI solutions. Papers exploring technical, processual, clinical and health policy support for the establishment of normative criteria to ascertain why/when/how cutting-edge AI solutions should be designed and be integrated into the standard of care are also sought.
Submissions that concentrate on responsible design perspectives and proffer insights into addressing these intricate concerns are particularly welcome. One key objective of this panel is to foster an in-depth/transdisciplinary dialogue on the development of AI technologies capable of potentializing expected/hoped-for outcomes in the health sector while adhering to rigorous ethical, legal, and societal demands.