Technological and industrial transformations driven by data science and artificial intelligence are profoundly impacting the field of materials science, presenting both unprecedented opportunities and significant challenges for the innovation of energetic materials. As an emerging technology, machine learning offers a novel research and development paradigm for the molecular design and synthesis of energetic materials and holds promise for overcoming long-standing bottlenecks such as low efficiency, high costs, and prolonged development cycles. Although some successful applications have been reported, the integration of machine learning across the full research cycle of energetic molecules—design, screening, synthesis, and performance validation—remains relatively underdeveloped compared to its application in other advanced materials domains. This review outlines the current landscape of machine learning-assisted development of energetic materials, with a focus on its applications in molecular design, single-property prediction, and multi-property simultaneous prediction. Nonetheless, the use of machine learning to guide the design and synthesis of energetic materials with targeted properties remains fraught with challenges. Future efforts should prioritize the establishment of data quality control and standardization frameworks, the development of interpretable machine learning models, and the creation of interdisciplinary integration platforms, all aimed at accelerating the efficient discovery of high-performance energetic materials.