Legal terms can be mapped from annotated corpuses. A "semantic annotation" presents a much more precise description of the knowledge contained in the texts and of its semantics in the legal field. A semantic annotation has to be well defined, easy to understand by the experts in the field, and not ambiguous. To meet these requirements, a semantic annotation must be based on a formal model of the domain. Texts with terms annotated with the corresponding legal concepts describing this field will improve the process of extracting information from texts and documents with legal content for Romanian as well. Texts will help achieve far more qualitative results in Automated Translations by disambiguating legal terms. Annotation of legal texts can not be achieved without specific tools. In this case study, we will proceed to analyzing the Constitution concept with the PaLinKa software. In recent years, the need to produce reusable cells has led to an increasing use of XML encoding in annotation. As a result, annotation can not be applied using simple text editors. In addition, discourse annotation is usually complicated and requires specialized tools. In this subchapter, we will present the most important features of an annotation speech tool. PaLinKa, the instrument presented in this paper, meets all these requirements, being appropriate for annotation of the legal texts.