This model was developed by scam.ai with a core belief: trust is the foundation of a strong society. With the rise of Generative AI, this foundation is under threat. Our mission is to protect people from scams and misinformation enabled by GenAI. We have a better state of arts detection algo available in our webiste scam.ai, feel free to have a try!

This model specializes in detecting the likelihood of deepfake faceswaps. It does not cover purely AI-generated images, as we consider faceswaps to pose the greatest risk due to their potential for misuse in scams. The system focuses on detecting a single face with the highest confidence score. If the image uses AI-powered super-resolution filters, the model’s accuracy may be affected, and it will warn you that the image has been AI-filtered.