Major domestic internet companies in China have been warned against the security risks of applying “deepfake” technology that could be used to create synthetic audio or video content of people speaking.
Chinese regulators have recently held talks with the companies on potential problems involving “deepfakes” to ensure healthy cyberspace and protect personal information.
“Deepfake” technology uses artificial intelligence to automatically generate, forge or tamper with images, audios, or videos, which are fake but appear highly realistic.
“Misuse of deepfakes can have serious consequences,” said Zuo Xiaodong, vice president of the China Information Security Research Institute.
Zuo said some people abroad have used face-swapping technology to fake images of artists, which has sparked heated debates concerning the technology.
The regulatory reminder is to sound alarms on the potential hazards and require the firms to timely report clues that may involve crimes, Zuo said.
“The talks have sent a clear signal that Chinese regulators will supervise the new technologies and applications of the internet companies in the early stages of their development, helping them plug security loopholes,” said Yin Sheng, an internet industry expert.
Yin suggested that the regulators should also improve the effectiveness and quality of supervision.
Cyberspace regulators will step up inspections to ensure that internet enterprises abide by laws and regulations and foster a healthy network ecosystem, according to the regulators.