Democratizing Truth: Why We Built a WhatsApp Deepfake Detector Anyone Can Use

A new deepfake detection service aims to counteract the imbalance between accessible AI creation tools and complex detection capabilities by offering a simple solution through WhatsApp. Users can text 218-NO-FAKES and send any suspicious media to receive an immediate analysis that not only confirms if the content is manipulated but also explains how it was likely created. Designed to eliminate barriers such as downloads, accounts, and specialized knowledge, the service analyzes images, videos, and audio with 94% accuracy. It has been tested in real-world scenarios, like identifying AI-generated receipts and synthetic voicemails, providing clarity for users confronted with potentially deceptive media. By making verification tools widely accessible, the initiative seeks to diminish the incentive to create malicious deepfakes, promoting a movement where the ability to discern truth is available to everyone.