Contributory Injustice, Epistemic Calcification and the Use of AI Systems in Healthcare

Abstract

AI systems have long been touted as a means to transform the healthcare system and improve service user outcomes. However, these claims frequently ignore the social context that leaves service users subject to epistemic oppression. This paper introduces the term “epistemic calcification” to describe how the use of AI systems leads to our epistemological systems becoming stuck in fixed frameworks for understanding the world. Epistemic calcification leads to contributory injustice as it reduces the ability of healthcare systems to meaningfully consider alternative understandings of people’s health experiences. By analysing examples of algorithmic prognosis and diagnosis, this paper demonstrates the challenges of addressing contributory injustice in AI systems and the need for contestability to focus on more than the AI system and on the underlying epistemologies of AI systems.