Manual Software requirements engineering has encountered significant challenges related to time overhead, high human effort, susceptibility to errors, and limited scalability for dynamic change handling, ranking and prioritization of non-functional requirements and requirements change requests. This study proposes an adaptive, dynamic, scalable, and AI-driven optimization model for software requirements scaling and prioritization using advanced machine learning and hybrid AI-based prioritization to improve accuracy and efficiency in requirement decision-making.The proposed approach employs hybrid AI-driven framework integrating machine learning models for requirement classification and prediction, natural language processing for text processing, optimization-based scoring for ranking and prioritization, and domain-aware AI models. Software Requirements Prioritization and Change Management Model (SRPCMM) embed AI-driven analysis across requirement engineering phases, enabling optimization-based scoring and dynamic re-prioritization, Natural Language Processing (NLP) -based ambiguity reduction, and efficient change management. A novel risk-adjusted weighted priority scoring mechanism supports realistic and integrated criteria-driven evaluation. The experimental evaluation on an industrial software dataset has shown a 25–30% improvement in prioritization accuracy. The proposed model addresses key limitations of traditional requirement engineering by enabling integrated, automated, and intelligent decision-making processes.