Yoti Strategy for Detecting Generative AI Threats

This technical report outlines Yoti's comprehensive strategy to address the threats posed by generative AI, particularly in the realms of identity verification and content integrity. The report details two primary attack vectors: presentation attacks and injection attacks, emphasizing the importance of early detection during the verification process. Yoti's approach includes the use of proprietary technologies such as MyFace® Live and MyFace® SICAP, which are designed to detect and mitigate these threats effectively. The report further discusses the potential risks associated with generative AI, including deepfakes, identity theft, and privacy concerns. It highlights the challenges that platforms face in moderating AI-generated content and the need for collaborative efforts to enhance security measures. Additionally, the document addresses the evolving nature of generative AI and the necessity for continuous advancements in detection technologies to combat these emerging threats.