This article aims to provide an overview of the application of machine translation technology in the field of legal translation, exploring its potential, challenges, and future directions. Firstly, it reviews the development process of machine translation technology, including rule-based methods, statistical machine translation, and the application of deep learning methods. Secondly, on the basis of in-depth deconstruction of the operating rules of the big language model, it is demonstrated that the highly modeled written legal language is highly consistent with the underlying logic of the big language model because of its standardization, accuracy and de-contextualization, so compared with other styles, machine translation technology will achieve better translation results in the field of legal translation. In addition, multimodal technology can also be applied to court interpretation, which greatly alleviates the shortage of qualified interpreters and maintains judicial justice. Furthermore, it discusses the human-machine collaborative model in legal translation, emphasizing the importance of human proofreading and review in ensuring translation accuracy and reliability. Lastly, it summarizes the prospects and challenges of machine translation technology in the field of legal translation. Through a systematic review and analysis of relevant literature, this article reveals the immense potential of machine translation technology in legal translation. It can significantly enhance the efficiency and quality of legal translation, thereby enhancing the capacity of legal language services.