Relative explainability and double standards in medical decision-making

Abstract The increased presence of medical AI in clinical use raises the ethical question which standard explainability is required for an acceptable and responsible implementation AI-based applications contexts. In this paper, we elaborate on emerging debate surrounding standards AI. For this, first distinguish several goods usually considered to contribute general, specific. Second, propose understand value relative other available norms explainable decision-making. Third, pointing out that accept heuristics uses bounded rationality decision-making by physicians, argue decisions should not be measured against idealized diagnostic process, but according practical considerations. We conclude, fourth, resolve issue explainability-standards relocating AI’s certifiability interpretability.