IA generativa y bienestar laboral: efectos en el desempeño percibido y la carga laboral en el sector público

Generative artificial intelligence is expanding across the public sector, yet how it affects the well-being of those who use it remains unknown. This study examined which antecedents of its adoption strengthen the emotional and psychological well-being of public servants and how well-being, in turn, shapes perceived performance and workload. Using PLS-SEM with bootstrap resampling, 433 public servants from El Oro, Ecuador, were analyzed. Three factors explained 79.5% of the variance in well-being: trust and security in AI was the strongest predictor (β = .338; p < .001), followed by perceived usefulness (β = .322; p < .001) and ethics and responsibility in use (β = .196; p < .001), whereas ease of use, institutional training, and fear of replacement were not significant. Well-being proved decisive, accounting for 76.0% of perceived performance (β = .872) and, jointly with it, 73.4% of workload reduction, operating mainly through performance (indirect effect β = .554). The importance-performance map analysis confirmed trust as the priority management lever. Adopting generative AI responsibly thus requires ethical governance, technological trust, and systematic well-being monitoring, redefining the priorities of public human resource management.