The rapid development of artificial intelligence in the financial sector has driven the shift from traditional investment advisory systems to data driven, automated models. However, current robo-advisor systems still rely primarily on static investor profiles and fixed portfolio optimization models, leading to significant limitations in contextual personalization and adaptation over time. In this context, Generative AI emerges as a novel approach with the ability to infer contextual information, integrate knowledge, and generate natural explanations for financial decisions. This research proposes a Generative AI framework for personalized asset management, aiming to integrate investor data, contextual modeling, and Generative AI inference mechanisms within a unified, multi-layered architecture. The proposed framework comprises six main components including investor data, investor profiling, context modeling, generative AI inference engine, recommendation and interpretation, and feedback and continuous learning mechanisms. This framework enables the transition from rule-based financial advice to an intelligent advisory system capable of real-time adaptation and personalization. In terms of contribution, the research expands the role of Generative AI from forecasting tools to financial inference systems, while proposing an integrated architecture for personalized asset management. Furthermore, the research utilizes the technology-organization-environment (TOE) framework to analyze the practical implementation conditions of the system, thereby clarifying the opportunities and challenges in applying Generative AI to the asset management field. The research findings provide both theoretical and practical value, contributing to the development of next-generation financial advisory systems that are highly explanatory, adaptable, and personalized.

Keywords: Generative AI, Wealth Management, Robo-advisory, AI for Financial Analysis, Personalized Investment.