---
id: "concept-algorithmic-override"
type: "concept"
source_timestamps: ["\\\"§ Sometimes", "people don’t want to know.\\\""]
tags: ["human-in-the-loop", "decision-making"]
related: ["claim-explanations-increase-override", "action-encourage-second-guessing"]
definition: "The act of a human operator rejecting or altering an AI system's recommendation, which occurs more frequently when the operator engages with the AI's underlying reasoning."
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"
---
# Algorithmic Override

**Definition:** The act of a human operator rejecting or altering an AI system's recommendation, which occurs more frequently when the operator engages with the AI's underlying reasoning.

Algorithmic override occurs when a human decision-maker chooses to reject or alter the recommendation provided by an AI system. The study found a **direct correlation between transparency and override rates**: when participants actually chose to view the AI's explanations, they were **approximately six percentage points more likely to challenge the AI's recommendation** (e.g., by approving both loans in the experimental setup).

This highlights the functional value of [[concept-explainable-ai]]: when engaged with, it successfully prompts human critical thinking and reduces blind compliance. It is the payoff that overcoming [[concept-willful-ignorance-in-ai]] delivers, and the behavioral target of [[action-encourage-second-guessing]]. See the claim [[claim-explanations-increase-override]].

**Enrichment note:** The working paper reports that when explanations revealed the AI had penalized non-White or female borrowers, participants were more likely to override the AI's *profit-maximizing* recommendation — the causal link (explanations → higher override of biased recommendations) is explicitly documented. **The precise "about six percentage points" figure is not directly verifiable from public summaries and should be treated as provisional.** This effect is the *pro-explanation upside* that a pure avoidance narrative under-states: in fairness-salient contexts, explanations demonstrably improve critical judgment.