This paper introduces an algorithm, termed self-adaptive genetic grey wolf optimizer (self-adaptive genetic GWO), for privacy preservation using a C-mixture factor. The C-mixture factor improves the privacy of data, in which the data does not satisfy the privacy constraints, such as l-diversity, m-privacy and k-anonymity. Experimentation is carried out using the adult dataset and the effectiveness of the proposed self-adaptive genetic GWO is checked depending on the information loss and the average equivalence class metric values and is evaluated to be the best when compared to other existing techniques with low information loss value as 0.1724 and average equivalence class value as 0.71, respectively.