63/100

Moderate evidence

ProblemPublished recordHealthProblem Space

Generative AI Use and Two-Wave Change in Well-Being and Ill-Being: A Pre- and Post-generative AI Era Panel Study

Truvace record
TRV-2026-3028raw data (JSON)
Version
v1Published
First published
Evidence type
Observational study
Sources
1 sourcePeer-reviewed
sha256
e8f254d0fe58…check it

As generative artificial intelligence (AI) rapidly spreads, concerns about its impact on users' mental health have surged. However, previous studies have yielded contradicting results owing to theoretical and methodological limitations. To address these gaps, this study investigates how generative AI adoption, usage time, and multidimensional AI dependency are related to two-wave changes in subjective well-being (self-esteem and life satisfaction) and ill-being (loneliness, anxiety, and depression). A two-wave l…

The problem

In a two-wave panel survey of 1,022 South Korean adults (October 2022 and March 2026), withdrawal-type dependency on generative AI, meaning distress or frustration when unable to use it, was associated with less decline in loneliness.

What this doesn’t fix

The depression association was described as suggestive but less robust, and the findings are associations from a survey rather than proof that dependency causes poorer mental-health change.

Sources

1 on file
  1. Peer-reviewed2026-11-17

Cite this record

v1

Truvace Impact Record TRV-2026-3028, v1: “Generative AI Use and Two-Wave Change in Well-Being and Ill-Being: A Pre- and Post-generative AI Era Panel Study.” Truvace, 2026-11-18. /record/TRV-2026-3028 (accessed on the day you copy it). sha256 e8f254d0fe58e4a1…

Change history

1 version

Every change made to this record since it was first published, and why. Nothing has changed since it went up.

  1. Publishedv1e8f254d0fe58…

    Certified into the record

Check this record hasn’t been altered

Check this record hasn’t been altered
How to check it yourself, without trusting this page

Download any version from /api/record/TRV-2026-3028 and hash its canonical field yourself, for example with shasum -a 256. You should get its content_hash. Each version’s text also ends with prev:and the hash of the version before it (version 1 points to 64 zeros). If anyone had changed a single character of any version after it was saved, those hashes wouldn’t line up.