<p>This study examines the double-edged impact of generative artificial intelligence (AI) appraisals on workplace cheating behavior (WCB) through the lens of the transactional theory of stress and coping and self-determination theory. Data were collected from 458 employees and their supervisors at two time points. Results indicate that challenge appraisals of generative AI enhance psychological empowerment, thereby reducing WCB, whereas hindrance appraisals of generative AI diminish psychological empowerment, leading to increased WCB. Furthermore, job complexity moderates these relationships: higher job complexity weakens the positive impact of generative AI challenge appraisals and strengthens the negative impact of generative AI hindrance appraisals on psychological empowerment. This study offers novel insights into the ethical implications of generative AI in the workplace and provides practical guidance for mitigating unethical behaviors through effective generative AI integration.</p>