The advent of Maritime Autonomous Surface Ships (MASS) brings transformative changes to shipping, accompanied by unprecedented safety and regulatory challenges. Traditional risk assessment frameworks, typically sequential or isolated, do not appear to be suitable for managing the complex, interconnected risks of autonomous systems. This paper proposes a parallel framework that combines System-Theoretic Process Analysis (STPA) and Fault Tree Analysis (FTA), enabling a holistic and dynamic risk assessment method tailored to the unique demands of MASS. STPA proactively identifies unsafe control actions and system vulnerabilities in the control structure, and FTA decomposes high-level risks into specific failure pathways, quantifying their likelihoods, and identifying their root causes. By running these methods in parallel, the framework supports iterative improvements and feedback between system and component-level risks, enhancing the efficiency and comprehensiveness of the risk assessment process. It enhances coverage of cascading failures, critical vulnerabilities, and mitigation strategies, in a manner that is both rigorous and comprehensible for designers, regulators, and stakeholders. A simplified example involving collision risk in a crossing situation is used to demonstrate the framework’s application. The approach supports regulatory alignment and provides a strong foundation for the development of robust safety standards in autonomous maritime operations.