---
id: "claim-ai-undermines-trust"
type: "claim"
source_timestamps: ["§ Warning Signs Ahead"]
tags: ["organizational-culture", "trust"]
related: ["framework-four-risks-ai-relationships", "quote-ai-sycophancy"]
confidence: "medium"
testable: true
speakers: ["Constance Noonan Hadley", "Sarah L. Wright"]
sources: ["adoption"]
sourceVaultSlug: "hbr-seg-adoption"
originDay: 9
articleStem: "hbr-sig-53-ai-personal-support-risky"
sourceUrl: "https://hbr.org/2026/05/employees-are-relying-on-ai-for-personal-support-thats-risky"
sourceTitle: "Employees Are Relying on AI for Personal Support. That’s Risky."
---
# AI Adoption Undermines Opportunities to Build Interpersonal Trust

**Claim:** AI disrupts the crucial exchanges of *giving and receiving help* that build interpersonal intimacy and mutual reliance at work.

**Evidence:** Citing a **May 2025 survey by MOO** in which **65%** of workers said they turn to AI tools before asking a colleague for assistance, the authors argue that because AI is faster and free of judgment, employees bypass their peers. By removing the need to go to colleagues for help, AI eliminates the micro-interactions that serve as a core antidote to loneliness — the feeling that others *"have your back."* As this dependence grows, organizational trust in peers and leaders is likely to deteriorate.

**Confidence:** Medium. **Testable:** Yes.

This is the third mechanism in [[framework-four-risks-ai-relationships]]. The appeal of judgment-free AI is voiced in [[quote-ai-sycophancy]].

**Enrichment context:** The descriptive part (more workers go to AI first) is empirically supported; Workday finds 16% have *less patience for small talk* since adopting AI. The trust-erosion inference is grounded in well-established theory that help-seeking and help-giving build interpersonal trust and psychological safety — but it is inferential and design-dependent, hence medium confidence. Counter-evidence exists: some studies show AI can *enable* new collaboration by coordinating tasks and surfacing expertise when deliberately designed to.