John and Melinda Thompson Director of AI in Medicine (Integration lead), Bioethicist
The Hospital for Sick Children
On the Inextricability of Explainability from Ethics: Explainable AI does not Ethical AI Make
Explainability is embedded into a plethora of legal, professional, and regulatory guidelines as it is often presumed that an ethical use of AI will require explainable algorithms. There is considerable controversy, however, as to whether post hoc explanations are computationally reliable, their value for decision-making, and the relational implications of their use in shared decision-making. This talk will explore the literature across these domains and argue that while post hoc explainability may be a reasonable technical goal, it should not be offered status as a moral standard by which AI use is judged to be ‘ethical.’
