The fundamental challenges identified in this research to making law computable are not addressed by developments in machine learning or even deep learning. Machine learning is primarily concerned with identifying patterns rather than thinking. Predictive models that improve over time as a result of exposure to fresh data are commonly found in machine-learning software. In order to suggest machine learning or deep learning as an option for responding to the issue of making law attainable, one must be able to specify the task that one wishes to have the machine-learning model complete, list the inputs and outputs of the program, and provide the training data that the software will use to create a predictive model. Furthermore, unlike human readers, machine-learning algorithms are incapable of reasoning or comprehending texts written in natural language. In legal AI, classical symbolic AI techniques still play a significant role, even though machine learning may prove beneficial for legal predictive analytics and categorization tasks like e-discovery document analysis. Users who use modern information retrieval systems are able to submit complete documents as search queries and conduct innovative, novel types of information searches. Numerous experts in the legal field will have to reinvent themselves and discover new specializations. In response to that disturbance, the State will need to rearrange itself and attempt to strike a balance. To maintain ethical, responsible, and legally sound procedures that respect human rights and the welfare of society while simultaneously utilizing AI’s potential benefits for the legal system is the fundamental challenge. This makes it necessary to do research in order to alter legal frameworks so that they can keep up with the rapidly changing field of artificial intelligence.