Trust Stamp, the Privacy-First Identity Company, announced that it has filed provisional patent #63/662,575 with the US Patent and Trademark Office for a new methodology to detect injection attacks in biometric authentication processes, including attacks executed using deep fake images and videos.
Injection attacks targeting biometric processes typically bypass the camera on a user’s device or inject video or still images captured in a different context into the data stream between the user’s device and the server to which they are authenticating.
Also Read: FriendliAI integrates with Weights & Biases to streamline Generative AI deployment workflows
Dr Norman Poh, Trust Stamp’s Chief Science Officer, commented, “We already have a number of liveness detection technologies implemented, but there are now billions of daily attacks being perpetrated with a growing number of injection attacks using genuine artifacts captured out of context as well as deep fake images and videos. When genuine artifacts are used out of context, they may be able to pass legacy liveness detection tests. With rapid advances in generative AI technology, we always have to be watchful for deep fakes that can defeat liveness tests. This latest presentation attack detection technology that we have patented targets injection attacks regardless of the artifacts being used.”
SOURCE: GlobeNewsWire