Work

Simulated quiet quitting risk in generative agents under crisis management strategies

Year: 2026
Type: article
Abstract: Quiet quitting refers to employees maintaining outward compliance while reducing psychological engagement and discretionary effort. Because such withdrawal is partly concealed, it is difficult to observe directly in conventional organizational research. This study develops an exploratory generative-agent simulation to examine how GPT-4o-mini produces compliance and withdrawal-related responses under different crisis-management prompts and personality archetypes. We implemented a factorial computational experiment crossing two managementstrategy prompts, control-oriented and commitment-oriented, with six predefined personality archetypes. The primary outcome was a narrow simulated proxy for quiet-quitting risk, defined as outward compliance accompanied by a model- generated turnover-intention label. The results show that management-strategy prompts and personality conditioning were associated with different patterns of generated compliance-with-withdrawal responses. However, these outputs should not be interpreted as measurements of real employee behavior or as empirical validation of organizational theories. Rather, they show how one language model, under specified prompt conditions, generates theoretically relevant synthetic response patterns that may inform future hypothesis development. Expert ratings and diagnostic analyses are presented as plausibility assessments of the generated text, not as validation against real employee data. The study contributes a transparent computational workflow for exploring compliance-with-withdrawal patterns while emphasizing the need for less leading prompt designs, independent outcome coding, and external validation using human-subject or organizational datasets.
Language: English
Open Access status: gold
Estimated APC paid: $1,520
OpenAlex ID: W7167714589
FWCI:
Cites: 43
Cited by:
Related to: 10