Against the backdrop of urban transformation focusing on stock renewal and quality improvement, old residential communities across China are plagued by aging landscape facilities, monotonous spatial functions, and poor environmental adaptability. Conventional landscape renovation methods, which rely heavily on manual design and empirical judgment, usually suffer from low construction efficiency, serious design homogenization, and insufficient consideration of residents’ actual living demands. With the rapid development and widespread application of artificial intelligence technology, innovative solutions have emerged to support refined, intelligent, and human-centered landscape upgrading of aging residential neighborhoods. Taking the landscape renewal of old residential areas as the core research theme, this study systematically sorts out the prevailing problems existing in current community landscape renovation practices. By integrating the technical strengths of artificial intelligence in multi-source data acquisition, intelligent quantitative analysis, scenario simulation, and dynamic operation management, this research proposes targeted renovation strategies from the perspectives of demand matching, spatial optimization, ecological improvement, and characteristic landscape shaping. The research findings aim to resolve the inherent drawbacks of traditional renovation modes, create high-quality and livable community landscape spaces that adapt to residents’ daily needs and urban development trends, and provide feasible theoretical references and practical guidance for the stock renewal of urban old residential communities.