PurposeArtificial intelligence (AI) radically transforms organizations, yet ethical AI’s effect on employee innovation remains understudied. Therefore, this study aims to explore whether responsible artificial intelligence (RAI) enhances high-tech employees’ innovative work behavior (IWB) through creative self-efficacy (CSE) and employee mental health and well-being (EMHWB). The study further examines how leaders’ RAI symbolization (LRAIS) moderates RAI’s effect.Design/methodology/approachThrough structural equation modeling, 441 responses of high-tech firms’ employees from Pakistan were utilized for hypotheses testing via SmartPLS-4.FindingsThe results revealed that second-order RAI enhances employees’ IWB. The effect was supported directly and indirectly through CSE and EMHWB. Findings also showed that LRAIS significantly moderates RAI’s influence on CSE, on the one hand, and EMHWB, on the other.Practical implicationsHigh-tech firms’ managers can fix AI-outlook issues that impair their employees’ IWB by prioritizing an ethical AI design involving actions like AI control mechanisms, bias checks and algorithmic audits. Similarly, these managers should facilitate RAI discussions and targeted trainings focusing on employees’ cognitive development and well-being. Likewise, RAI embracement programs and evaluations for leadership positions could be incorporated into high-tech firms.Originality/valueThis study advances the mainstream AI literature and addresses a notable gap concerning RAI’s influence on employees’ IWB while grounding in social cognitive theory. Moreover, this study unveils how CSE and EMHWB affect IWB within RAI milieus. Additionally, through signaling theory, it underscores the significance of LRAIS in amplifying the direct association between RAI, CSE, and EMHWB within high-tech firms in emerging markets.