In recent years, there has been a noticeable increase in reliance on social media news and online news portal platforms for keeping up with world events and checking their authenticity. While online platforms provide the convenience of obtaining news at any time and from any location, it has also become a fertile environment for the distribution of erroneous information, resulting in the widespread spread of fake news and negative societal effects. To avoid user misunderstandings and promote positive societal advancement, it is critical to distinguish between authentic and falsified news posted on social media. This research work extended the state-of-the-art COVAX reality dataset by adding two more features a news body and a news summary, which led to the stance detection of news. Moreover, this research work proposes an innovative approach to authenticating the news content using three different text summarization techniques (LEAD-3, sequence-to-sequence (Seq2Seq), and Bidirectional Encoder Representations from Transformers (BERT)). Among the summarization models we examined, the BERT model consistently outperformed others, delivering the best precision value 89.47%, Recall 91.64%, and F-measure 90.54% and 0.8335 BERT score. Moreover, this study highlights the potential of utilizing the transformer-based model, ELECTRA Large, to address the dissemination of fake news, achieving an impressive accuracy of 88.97%.