---
id: "claim-financial-incentives-dampen-transparency"
type: "claim"
source_timestamps: ["\\\"§ Sometimes", "people don’t want to know.\\\""]
tags: ["incentive-design", "behavioral-economics"]
related: ["concept-willful-ignorance-in-ai", "contrarian-transparency-desire"]
confidence: "high"
testable: true
speakers: ["Alex Chan"]
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"
---
# Financial incentives decrease the desire for AI transparency

**Confidence:** high · **Testable:** yes · **Attributed to:** [[entity-alex-chan|Alex Chan]]

When a user's compensation is directly tied to the outcome of an AI-assisted decision, their desire to understand the AI's reasoning decreases. In the study, **participants whose bonuses depended on loan repayments were nearly 20% more likely to decline viewing AI explanations compared to participants receiving a flat fee.**

This proves that performance-based financial incentives actively compete with and suppress the pursuit of algorithmic transparency — the financial engine behind [[concept-willful-ignorance-in-ai]] and evidence for the [[contrarian-transparency-desire|contrarian claim]] that people do not naturally want transparency. It is the empirical basis for the practitioner move in [[action-align-incentives-critical-engagement]] and the still-open [[question-optimal-incentive-structures]].

**Enrichment note:** The direction and mechanism (performance-based pay dampens explanation demand) are clearly validated across Chan's NBER/HBS paper, the D³ article, and Meyer's synopsis — when bonuses depend on repayment, participants seek predictions but avoid explanations, dodging information that could force a choice between personal benefit and fairness. The D³ article's comparable figure is that lender-aligned participants were *about 10 percentage points* more likely to skip explanations than neutrally paid participants. **The exact "nearly 20%" magnitude is not articulated in accessible public sources and should be treated as provisional pending direct inspection of the full paper tables.**


## Related across articles
- [[claim-input-metrics-punish-efficiency]]
- [[contrarian-metric-penalties]]