High-stakes clinical decisions are increasingly shaped by algorithmic systems that may inform diagnosis, treatment, triage, discharge, risk classification, and access to scarce interventions. Ethical discussion has often emphasized notice, disclosure, transparency, and explanation. These protections are important but incomplete when an affected patient cannot ask why a consequential decision was made, obtain reconsideration by an empowered clinician, or secure revision when the decision cannot be adequately defended. This article develops a rights-based normative argument for a proposed right to a human reason. The right is not presented as an existing universal legal entitlement, nor does it presume that human judgment is inherently more accurate than artificial intelligence. Rather, it protects the patient’s standing as someone to whom consequential clinical decisions should be answerable. A human reason is defined as an intelligible, patient-directed justification that an appropriately competent and authorized human agent can assess, own, defend, revise, or withdraw. The article distinguishes explanation from justification and justification from contestability; proposes proportional triggers for the right; and specifies minimum procedural safeguards from notice through appeal. It also develops a risk-sensitive model for high-stakes decisions while recognizing emergency, resource, expertise, equity, and institutional constraints. The proposal remains normative and requires empirical testing of accessibility, burden, delay, error correction, reviewer independence, equity, and unintended effects before any implementation claim would be warranted.
