---
id: "concept-moral-quandary-avoidance"
type: "concept"
source_timestamps: ["¶5", "\\\"§ Sometimes", "people don’t want to know.\\\""]
tags: ["ethics", "behavioral-psychology", "algorithmic-bias"]
related: ["concept-willful-ignorance-in-ai", "claim-bias-suspicion-increases-avoidance"]
definition: "The behavioral tendency to actively avoid information that might reveal ethical issues (like algorithmic bias), thereby preventing cognitive dissonance or the need to alter a profitable decision."
sources: ["adoption"]
sourceVaultSlug: "hbr-seg-adoption"
originDay: 9
articleStem: "hbr-edu-37-employees-not-questioning-ai"
sourceUrl: "https://hbr.org/2026/06/employees-arent-questioning-ai-advice-enough"
sourceTitle: "Employees Aren’t Questioning AI Advice Enough"
---
# Moral Quandary Avoidance

**Definition:** The behavioral tendency to actively avoid information that might reveal ethical issues (like algorithmic bias), thereby preventing cognitive dissonance or the need to alter a profitable decision.

In decision-making environments, individuals actively avoid information that might place them in a moral quandary. In [[entity-alex-chan|Chan]]'s loan-approval experiment, when participants were informed that viewing an AI's explanation **might reveal that race or gender influenced the algorithm's recommendation, their rate of avoiding the explanation rose by more than 10 percentage points (reaching 23%)**.

People prefer to remain ignorant rather than confront evidence of bias, because knowing about the bias would force them to meaningfully change their decisions or experience cognitive dissonance and moral discomfort while following the AI's advice. This is a specific, ethics-driven variant of [[concept-willful-ignorance-in-ai]], and it depends on the reader understanding [[prereq-algorithmic-bias]].

See the supporting evidence in [[claim-bias-suspicion-increases-avoidance]].

**Enrichment note:** Chan's D³/HBS article describes a condition where fairness auditing was made *salient*, and lender-aligned participants were about 10 percentage points more likely to skip explanations than neutrally paid participants — directionally and qualitatively confirming the mechanism. **The exact figure "more than 10 percentage points (reaching 23%)" is not independently visible in public summaries and should be treated as provisional pending the full working-paper tables.** The framing maps directly onto the information-avoidance literature (Golman, Hagmann & Loewenstein 2017) on avoiding morally obligating information.