Law can be defined as a tool for resolving human conflicts. Its proper application depends on the ability of the involved agents to understand texts laden with its inherent semantics, typical of the field. The use of explainable artificial intelligence in this task still requires further study. This work aims to initiate research focused on the application of explainable AI within the scope of Brazilian legislation. We hope that with this model, we can interpret these texts and explain the results of the algorithm's predictions by understanding the ‘why yes?’ and ‘why not?’ To achieve this, we applied the SHAP technique to a set of legal texts in the Portuguese language. As part of the model evaluation, we will present the preliminary results to legal professionals for refinement. This model will serve as a basis for future work, where we will apply other techniques such as LIME and their combinations.