---
id: "concept-sycophantic-ai"
type: "concept"
source_timestamps: ["¶17"]
tags: ["ai-alignment", "critical-thinking", "edge-cases"]
related: ["concept-ai-persona", "concept-dark-triad-ai"]
definition: "An AI persona that is excessively agreeable and eager to please, which degrades work quality by dulling user critical thinking and validating weak ideas."
sources: ["tail1"]
sourceVaultSlug: "hbr-seg-tail1"
originDay: 1
articleStem: "hbr-tail-113-ai-personality-problem"
sourceUrl: "https://hbr.org/2026/06/does-your-ai-have-a-personality-problem"
sourceTitle: "Does Your AI Have a Personality Problem?"
---
# Sycophantic AI

While the [[concept-dark-triad-ai|dark triad AI]] represents *overt* hostility, toxicity in AI personas can take multiple forms. A **sycophantic AI** is endlessly agreeable, overly deferential, and excessively eager to please the user.

The authors note this interaction style can be **equally corrosive** to organizational performance, because it **dulls the user's critical thinking** and rubber-stamps weak, flawed, or unoriginal ideas. (This maps onto the AI-alignment community's concern with *sycophancy* — models mirroring a user's stated beliefs even when wrong.)

The key implication: optimizing an [[concept-ai-persona|AI persona]] is **not** simply about making the system 'warmer' or more polite. It is about designing an interaction style that fosters rigorous, high-quality collaboration — the balance-of-deference-and-pushback problem raised in [[question-optimal-persona-matching]].