Persona is hosting a 12-week internship for University at Buffalo PhD student Zhaofeng Si focused on detecting AI-generated synthetic selfies. The project aims to build a benchmark and taxonomy for selfie fraud. This initiative builds on Persona’s broader computer vision program, which recently produced a system for layout-aware representation learning across more than 10,000 ID document formats. That work, accepted to a CVPR 2026 workshop, targets open-set ID fraud discovery by learning patterns of authentic documents rather than only classifying known forgeries. If successfully incorporated into its platform, these capabilities could enhance Persona’s ability to detect sophisticated fraud attempts at scale and strengthen its competitive moat in the digital identity and anti-fraud market. Improved fraud detection may also support stronger positioning with security-sensitive customers and regulated industries. 

Source: https://www.tipranks.com/news/private-companies/persona-weekly-recap-9