Malicious visual media detection service is emerging, and many approaches have been proposed. However it is only in the early stages toward a robust, practical solution. The proposed solution, MANFID, if successful, will provide this emerging need with an effective tool to distinguish GAN generated image from real images.