Deepfakes are facing audiovisual material that has been digitally manipulated or manufactured synthetically using these approaches. DL is a very effective and adaptable technology that has found extensive applications in a variety of disciplines such as machine vision, computer vision, and NLP. DL has made an incredible. We conduct an in-depth analysis of the different methods employed and their application in identifying deepfake films or images. This work will be extremely helpful to scholars in this field as it contains the latest state-of-the-art methods for identifying deepfakes on social media and other content platforms. In addition, it contains a complete description of the latest methodology and datasets used, making it possible to compare with existing work. Researchers can use this comprehensive study to gain a deeper insight into deepfakes and try out new methods of fighting their harmful effects on society.