---
id: "question-healthy-ai-relationships"
type: "open-question"
source_timestamps: ["¶7"]
tags: ["ethics", "societal-impact"]
related: ["concept-emotional-support-ai", "quote-intimate-algorithms", "claim-therapy-top-use-case"]
resolutionPath: "Longitudinal psychological studies on users who rely heavily on AI for therapy and companionship, measuring emotional resilience, empathy, and human-to-human relationship quality over time."
speakers: ["Marc Zao-Sanders"]
sources: ["execution"]
sourceVaultSlug: "hbr-seg-execution"
originDay: 8
articleStem: "hbr-cl-77-new-data-using-ai"
sourceUrl: "https://hbr.org/2026/06/new-data-on-how-were-really-using-ai"
sourceTitle: "New Data on How We’re Really Using AI"
---
# Is it healthy for algorithms to manage intimate relationships?

**Open question:** Is it healthy — or desirable — for opaque algorithms to increasingly manage and influence our most intimate relationships?

Raised explicitly by [[entity-marc-zao-sanders]] (see [[quote-intimate-algorithms]]). Because therapy/companionship remains the top use case ([[claim-therapy-top-use-case]] / [[concept-emotional-support-ai]]), society is effectively running an **unmonitored psychological experiment** at scale.

**Resolution path:** Longitudinal psychological studies of heavy users of AI therapy and companionship — measuring emotional resilience, empathy, and human-to-human relationship quality over time — plus deeper sociological research into the effects of substituting human empathy with algorithmic text generation. Adjacent work on Replika-style emotional bonds and clinical iCBT/Woebot trials offers a starting evidence base and a set of guardrail proposals (disclaimers, crisis escalation, regular evaluation).