aceModerate evidence
More
Positive AI feedback raised dependency in a trial of 294 Japanese office workers
Study title: How workplace AI feedback shapes AI dependency: A randomized controlled trial on the moderating role of workplace social support among Japanese office workers
- Evidence type
- Randomized controlled trial
- Sample
- n = 294
- Sources
- 1 sourcePeer-reviewed
- Published
- 2 min read
- Truvace record
- TRV-2026-2566cite or check it
Study title: How workplace AI feedback shapes AI dependency: A randomized controlled trial on the moderating role of workplace social support among Japanese office workers
A five-day randomized controlled trial with 294 Japanese office workers tested how an AI chatbot's feedback shapes AI dependency, defined as habitual, automatic AI use with diminished autonomy. Participants held daily work-reflection sessions with the chatbot, which gave positive feedback on strengths, negative feedback on areas for improvement, or no work-related feedback in the control condition.
Read the study abstract
AI dependency—habitual, automatic AI use accompanied by diminished autonomy—is a growing concern in the workplace, yet the conditions under which it develops remain poorly understood. Two competing pathways have been proposed. Users with limited social support from colleagues may compensate for this deficit by depending on AI (the compensatory pathway ), or users with abundant support may be better positioned to leverage AI interaction, thereby fostering dependency (the enhancement pathway ). To test these pathw…
The problem
In a randomized trial of 294 Japanese office workers, positive chatbot feedback increased AI dependency one week later regardless of workplace support, and negative feedback did so among workers with abundant social support.
In brief
Dependency was measured one week after the final session. Positive feedback raised it regardless of how much social support workers had from colleagues. Negative feedback did not raise it overall, but did among workers with abundant workplace support.
The pattern matches the enhancement pathway, in which well-supported workers are better placed to build dependency on AI, rather than the compensatory pathway, in which isolated workers lean on AI to fill a gap. The authors conclude that dependency depends on the interaction between AI feedback and existing workplace relationships, not on AI characteristics alone.
Main points
- The trial enrolled 294 Japanese office workers who completed work-reflection sessions with an AI chatbot over 5 days.
- Three conditions were compared: positive feedback emphasizing strengths, negative feedback highlighting areas for improvement, and no work-related feedback.
- AI dependency was measured one week after the final session.
- Positive feedback increased dependency regardless of workplace social support levels.
- Negative feedback increased dependency only among participants with abundant social support, consistent with the enhancement pathway.
- Two competing explanations were tested: a compensatory pathway for workers lacking support and an enhancement pathway for well-supported workers.
The full story
Researchers define AI dependency as habitual, automatic AI use accompanied by diminished autonomy, and describe it as a growing workplace concern whose causes are poorly understood. They tested two proposed routes: workers with little colleague support compensating by relying on AI, or workers with plenty of support being better positioned to build dependency.
In the trial, workers had daily work-reflection sessions with a chatbot for five days. The chatbot gave positive feedback, negative feedback or none, and dependency was assessed a week after the last session.
Positive feedback worked as a general route to dependency across support levels. Negative feedback had an effect only for workers with abundant social support, which fits the enhancement pathway and not the compensatory one.
Sources
1 on file- Computers in Human Behavior: Artificial Humansdoi:10.1016/j.chbah.2026.100365
Talk it over
Who carries this cost, and what would have to change to prevent it?
Positive AI feedback raised dependency in a trial of 294 Japanese office workershttps://truvace.com/item/how-workplace-ai-feedback-shapes-ai-dependency-a-randomized-controlled-trial-on-
Who carries this cost, and what would have to change to prevent it?
The problem In a randomized trial of 294 Japanese office workers, positive chatbot feedback increased AI dependency one week later regardless of workplace support, and negative feedback did so among workers with abundant social support.
Answer in the threadHow do we know? What can this kind of evidence show, and what can’t it?
The evidence This rests on a randomized controlled trial with a sample of 294, from Computers in Human Behavior: Artificial Humans.
Answer in the threadWhat does it leave open? What else would you want to know before acting on it?
The record’s own caveat The outcome was short-term and self-reported, the analysis was exploratory and not preregistered, the key interaction was modest in size, and the crowdsourced Japanese sample limits generalizability.
Answer in the threadHave you seen AI used this way where you study, work or live? Did it go the same way?
Answer in the thread
The questions are the same on every record. The details under them are taken from this record as written.
What readers say