# Epistemic Dependence **Entity class:** Human-AI learning and judgment condition **Domain:** Artificial intelligence / education / public epistemology **Maturity:** Emerging research concept ## Definition **Epistemic dependence** occurs when people rely on artificial-intelligence systems for explanation, judgment or causal interpretation in ways that may preserve, supplement or displace the intellectual work through which independent understanding develops. ## Mechanism Dependence becomes hazardous when users cannot evaluate the system's evidence, confidence or causal attribution and when fluent answers substitute for the cognitive effort required to form and revise judgment. Calibrated uncertainty and visible evidence can instead make dependence more inspectable. ## Article Context [[articles/Epistemic Invariance in Conversational AI|Epistemic Invariance in Conversational AI]] connects conversational AI to public scientific literacy, automation bias and the risk that repeated accommodation teaches users to treat confidence, identity, tradition or cultural acceptance as substitutes for evidence. ## Evidence Boundary Reliance is not automatically displacement. The relevant question is whether AI use preserves the user's ability to understand evidence, detect error, revise beliefs and act without ceremonial deference to the machine. ## Relationships - [[wiki/Automation Bias|Automation Bias]] - [[wiki/Public Epistemic Capacity|Public Epistemic Capacity]] - [[wiki/Citizen AI Literacy|Citizen AI Literacy]] - [[wiki/AI Sycophancy|AI Sycophancy]] - [[wiki/Public AI Trust|Public AI Trust]] - [[wiki/Machine Epistemology|Machine Epistemology]] ## Sources / Provenance - [[articles/Epistemic Invariance in Conversational AI|Epistemic Invariance in Conversational AI]] — public-literacy argument and cited learning/automation research.