Abstract: Driven by digital and intelligent technologies, the balanced development and thorough penetration of financial services are of significant importance during the 15th Five-Year Plan period for accelerating the construction of a financial powerhouse and ensuring that the achievements of modernization benefit all people more extensively and fairly. The application of large AI models in inclusive finance scenarios substantially enhances resource accessibility for financially underserved groups, yet simultaneously amplifies the impact of ethical risks on the development and governance of inclusive finance. Considering both technical mechanisms and practical conditions, the heterogeneity between the operational logic of large AI models and that of inclusive finance, along with the complexity of application scenarios, leads to the emergence, spillover, and diffusion of ethical risks throughout processes such as data learning, algorithm operation, and knowledge generation. This creates tensions with value goals such as equity, security, stability, and sustainability. Consequently, it is essential to adopt a value alignment approach to prevent and control ethical risks in intelligent systems. It is necessary to fully elaborate the value alignment logic of large AI models across principles, intentions, behaviors, and incentives, guiding their operational rules, technical pathways, functional realization, and governance systems to align with the high-quality development requirements of inclusive finance. In terms of legal pathways, emphasis should be placed on effectively integrating the core tenets of inclusive finance into the formulation of norms governing large AI models, establishing a full-cycle risk control mechanism for identifying, intervening in mitigating ethical risks, smoothing ethical risks, and fostering a multi-stakeholder collaborative governance mechanism for value alignment through effective incentive compatibility of rights and responsibilities.