Implementing good governance strategies for human-centered AI in healthcare: connecting norms and context

Abstract

Human-Centered Artificial Intelligence (HCAI) in healthcare demands pairing at least two governance strategies: (1) A normative approach that ensures we protect the interests of healthcare actors at a regulatory level and (2) A contextual approach that allows understanding, responding, and adapting to such interests at a practical level. The first minimally requires that regulators ensure that AI innovations in healthcare are sufficiently safe to be implemented, and therefore respect some key overarching normative principles like objective performance and accountability. The second speaks to the capacity for the AI community to design, develop and implement AI innovations that respond to a variety of interests in practice and generate use value. These approaches both speak to the need for AI to aim at valuable ends; to be proven safe, effective, and fit for purpose; and to be designed for use in real settings. This chapter explains these approaches, their interrelation, and some of their practical and theoretical implications.

More information

Identifiers

ORCID for Michael Da Silva: orcid.org/0000-0002-7021-9847

Catalogue record

Export record

Altmetrics