1st March 2024, London, UK – Digital identity company Yoti has published its first Generative AI white paper, detailing its approach to combating deepfakes. Yoti has focussed their strategy on early detection, using tools to prevent AI-generated content or attacks at the point of source.
Whilst offering numerous benefits and advancements, generative AI also brings potential threats and challenges, including:
- Identity theft and fraud; the technology could be used to create synthetic identities and realistic forged documents.
- Fake content; the creation of highly convincing content, including images, videos and audio recordings.
- Privacy concerns; the ability to generate realistic images and videos raises privacy concerns, especially when it comes to creating fake content that features individuals without their consent.
Yoti has taken a layered approach to detecting and preventing deepfakes. They currently target two attack vectors: presentation attacks (direct) and injection attacks (indirect), with a focus on early detection during the verification process:
Liveness technology – Yoti’s proprietary passive liveness detection technology, MyFace, is compliant with iBeta ISO PAD Level 2. It performs passive liveness and detects presentation attacks (be those images presented on a screen, a printed image or 3D masks).
- SICAP – Yoti has developed SICAP, short for Secure Image Capture, a proprietary technology that works alongside Yoti’s liveness detection capabilities. SICAP can identify and prevent sophisticated injection attacks, ensuring that the images captured during a verification process are genuine and remain untampered with.
Paco Garcia, CTO at Yoti said, “As the threat of generative AI in identity verification accelerates, we have developed a comprehensive strategy focused on early detection. We are committed to developing leading technology which is at the forefront of combating evolving threats in the generative AI landscape. The combination of our liveness technology and SICAP solution gives businesses enhanced security and defence against fraudulent attempts and deepfakes.”
