Purpose The rapid integration of artificial intelligence (AI) into digital financial systems has created opportunities for innovation while simultaneously generating complex governance challenges. Although AI improves efficiency and financial inclusion, it also introduces risks related to cybersecurity, data governance, institutional readiness and regulatory compliance. This study aims to identify and prioritize AI-related risks in digital financial systems and examine their implications for AI governance, particularly in the context of Islamic digital finance. Design/methodology/approach This study employs a quantitative approach based on a survey of 260 respondents to evaluate perceptions of AI-related risks in digital financial environments. The Failure Mode and Effects Analysis (FMEA) framework is used to assess risks across four dimensions: potential occurrence, frequency, impact and detection capability. Risk Priority Numbers are applied to rank and classify risks, complemented by multilevel analysis at both category and item levels. Findings The findings indicate that the most significant risks are primarily governance-related rather than technological. Risk management capacity, human capital constraints and Sharia compliance emerge as dominant dimensions shaping AI risk governance. The analysis also shows that AI risk structures are nonuniform. While aggregate-level analysis identifies governance-related risks as dominant, item-level analysis reveals concentrated vulnerabilities, particularly cybersecurity risks, that may remain obscured within broader classifications. The findings further suggest the presence of a governance execution gap, where institutional awareness of AI-related risks is not always followed by effective mitigation practices. Research limitations/implications This study relies on perception-based survey data, which reflects respondents’ assessments of AI-related risks rather than direct observations of technological failures in real financial systems. Future research could extend the analysis by using case studies or institutional data from financial organizations that implement AI technologies. Practical implications The findings highlight the importance of strengthening institutional capacity through risk management frameworks, training and organizational readiness. Policymakers and financial institutions should adopt governance approaches that integrate technological safeguards with regulatory and ethical oversight, particularly within Islamic financial systems. Social implications Effective AI governance can strengthen public trust, support financial inclusion, and enhance the stability of digital financial systems. In Islamic financial contexts, alignment between AI systems and ethical as well as Sharia principles remains essential for maintaining institutional legitimacy and societal acceptance. Originality/value This study contributes to the literature by applying FMEA to AI risk governance in digital financial systems and by introducing a multilevel risk assessment perspective. The study further advances understanding of AI governance by identifying governance execution gaps and emphasizing the need to integrate technological and Sharia-based governance frameworks.
Darmawan Darmawan (Mon,) studied this question.