AI deepfakes that simulate dead people risk ‘haunting’ relatives, researchers warn
IndependentMAY 13 2024
Sexually explicit "deepfake" images, including those of children and generated using artificial intelligence, are flooding South Korea's chat groups, worrying authorities.The Korea Communications Standards Commission on Wednesday asked messaging app Telegram to delete and block sexually explicit deep fake images — created using artificial intelligence (AI) — from their platforms.
Hundreds of schools and universities, whose students had allegedly become victims — though unconfirmed — of deepfake sex crimes, were named Sunday on X.This comes after South Korean Pres. Yoon Suk Yeol called for a probe into rising deepfake abuse, and morphed photos of women — including students, teachers, and military personnel — were reportedly found in one Telegram chatroom with around 220K members.
South Korean media reported one Telegram chatroom with around 220K members, sharing morphed images of students, teachers, and military personnel.According to the National Police Agency, 180 digital sex crimes were registered in 2023 and 75.8% of the 120 suspects punished were teenagers. The Agency said it will make a seven-month push to tackle online sex crimes.
South Korea's deepfake pornography crisis has exposed the dark underbelly of the nation's digital landscape, where social media and messaging platforms' strong encryption and non-cooperation with law enforcement have inadvertently created a haven for cybercriminals engaging in digital sex crimes.
The potential for misuse aside, AI-powered deepfake technology is simply another tool for creative expression., Likeand withlike many other areasmediums, the ethical line can easily be drawn at consent and intent. The focus must now be on establishing norms, laws, and standards governing their use, balancing innovation with individual rights and public trust.