Employees who use ChatGPT and similar large language models may face lower trust from managers, according to a six-study paper published in The Leadership Quarterly. Researchers from the Hebrew University of Jerusalem and WHU Otto Beisheim School of Management found that supervisors trusted independently completed work 68 percent of the time but only 24 percent of the time when AI was involved. The trust drop was linked to poorer performance ratings, lower views of organizational citizenship, and reduced promotability. Clear company policies that encourage LLM use removed the statistically significant penalty, while bans or silence left it intact. A field study with academic supervisors produced similar results. The authors warn that fear of judgement could drive shadow AI use and recommend explicit guidance for both employees and managers.
Employees who use ChatGPT and similar large language models may face lower trust from managers, according to a six-study paper published in The Leadership Quarterly. Researchers from the Hebrew University of Jerusalem and WHU Otto Beisheim School of Management found that supervisors trusted independently completed work 68 percent of the time but only 24 percent of the time when AI was involved. The trust drop was linked to poorer performance ratings, lower views of organizational citizenship, and reduced promotability. Clear company policies that encourage LLM use removed the statistically significant penalty, while bans or silence left it intact. A field study with academic supervisors produced similar results. The authors warn that fear of judgement could drive shadow AI use and recommend explicit guidance for both employees and managers.