As digital system grow more complex and threats continue to evolve, adopting an Artificial Intelligence(AI)-driven approach is essential to improve the efficiency and accuracy of identifying security threats. This is especially important in regulated industries like healthcare, finance, and automotive, where protecting sensitive data and complying with regulations is crucial. Our research aims to recommend security requirements by proposing an adaptive framework that uses AI tools for asset identification, threat mapping, and requirements generation. We explore how key system assets, such as databases, authentication modules, and API gateways, can be mapped to both current and emerging threats to generate security requirements. Our approach utilizes advanced AI methods, including Large Language Models(LLMs) and Natural Language Processing(NLP) to identify critical assets from software artifacts and retrieve relevant threats for each asset. We present an initial end-to-end workflow model and demonstrate its feasibility using a sample case study. The approach’s effectiveness depends on the quality of input data and the clarity of its outputs, which can be challenging to achieve in complex environments. Nevertheless, we expect it to provide a solid foundation for enhancing the security requirements process.