Patronizing and condescending language can be represented positively or negatively some times. Patronizing can mean “giving support to”. Condescending can mean “superior attitude towards others”. The proposed work takes the news articles from various countries, and detect whether the news article contains any patronizing and condescending language or not. We apply deep learning (DL) techniques to address this issue, approaching it like a typical text classification task. We suggest using a pre-trained model. Distil-Bert and Data Augmentation. The proposed model DistilBERT achieved a precision of 76.18 and f1 score of 72.48. Data Augmentation with Synonym replacement achieved a precision of 78.12 and f1 score of 73.87.