Resemble AI

Resemble AI is a platform for generative AI security focused on voice, image, and video. It is designed for enterprise use, available either on-premises or via the cloud, and built with governance and compliance in mind. The platform offers three core capabilities: generate, verify, and detect. The generate function allows users to synthesize or clone a voice, with watermarks applied at the moment of creation before the content leaves the user’s infrastructure. The verify function protects audio, image, and video files with watermarks described as permanent, indestructible, and invisible, which travel with the file. The detect function identifies deepfakes across audio, image, and video, and is described as having zero-day model coverage, having been battle-tested against over 160 generative AI models. The platform also provides a multimodal deepfake detection feature that accepts an audio, image, or video file and returns in-depth analysis explaining the results. Resemble AI’s own voice AI models, such as Chatterbox, were built with provenance and detection in mind from the start.