Optimum use of natural resources and sustainable development are critical and challenging issues in today's world. One of the important economic sectors that is highly dependent on natural resources is the mining sector. Exploitation of mines as one of the important factors of economic growth and development of different regions has always attracted the attention of many researchers and policy makers. In the meantime, one of the major challenges is obtaining a social license (SL) to exploit mines, which can have profound effects on the process of extracting and using mineral resources. SL to Operate means obtaining the consent and support of the local community and various stakeholders to carry out mining activities. This license is not only limited to obtaining consent through govemment laws and regulations, but also includes gaining the trust and cooperation of the local community, non-governmental organisations and other relevant stakeholders. Failure to obtain a social permit can lead to delays in mining projects, increased costs, and even a complete stop of mining operations. The study employs machine learning algorithms to analyse extensive mining operations and community interaction data sets, including support vector machine (SVM), and random forests (RF). These algorithms not only facilitate the identification of underlying patterns but also enable the prediction of potential future impacts, thereby providing a robust analytical tool for stakeholders. These algorithms not only facilitate the identification of underlying patterns but also enable the prediction of potential future impacts, thereby providing a robust analytical tool for stakeholders. This research tries to provide a predictive model by analysing available data and identifying influential patterns that can help mining companies in obtaining SL and improving extractive processes. The RF demonstrates better performance compared to the SVM, with an accuracy of 88 per cent and an F1 score of 86 per cent, based on the evaluation metrics. Therefore, RF is a more suitable option for predicting SL acquisition.