SNU Implements Anti-Deepfake Technology for Graduation Photos

• •SNU has embedded anti-deepfake technology in 5,800 graduation portraits.
• •The initiative follows a notorious deepfake abuse case involving SNU alumni.
• •StealCut Protect technology disrupts deepfake generation by embedding invisible signals.

Seoul National University (SNU) has embedded anti-deepfake signals into 5,800 graduation portraits to prevent misuse, following a digital sex crime case involving manipulated photos of women. The university partnered with StealCut, a startup founded by alumni, to utilize its StealCut Protect technology, marking the first application of such technology in a South Korean university graduation album. The technology embeds nearly invisible pixel-level signals into images, disrupting deepfake models' ability to accurately reproduce facial features. This initiative aims to protect graduates from potential deepfake abuse, especially in light of the 'SNU Nth Room' case, which involved the creation and distribution of explicit deepfakes using graduation photos. The protective measures will also extend to the 2027 graduation album. StealCut provided the technology to SNU at no cost, emphasizing the importance of safeguarding personal images from digital manipulation.