Xinhua Addresses AI Voice Cloning and Biometric Security Governance

Xinhua News Agency reported that fraud prevention advice urging people to stay silent for three seconds when answering unfamiliar calls circulated widely on social media platforms. The alert warns that criminals may deliberately stay quiet to prompt responses, using brief speech samples to replicate vocal prints through artificial intelligence to impersonate victims. While staying silent provides an easy personal tip, the development of artificial intelligence technology has lowered the threshold for fabricating voices, and individuals cannot genuinely remain quiet online given wide exposure across voice messages, short videos, public speeches, and online conferences. Sensitive personal biometric data such as voice prints, fingerprints, faces, and irises face ongoing problems including excessive collection, induced authorization, and covert retention. Reported abuses include using artificial intelligence face-swapping to bypass facial recognition and drain bank cards, cloning celebrity voices for marketing, and simulating voices of relatives to defraud the elderly. Governance must move forward proactively by severely punishing illegal biological data collection, cracking down on illicit data markets, enforcing mandatory identification labels on generated synthetic content, establishing joint defense mechanisms against abnormal voices, and implementing biometric cross-verification during key transactions like transfers and remittances.

Reported illegal activities involving the misuse of biometric data include using artificial intelligence face-swapping to crack facial recognition systems and swipe victims' bank cards, cloning celebrity voices to drive sales traffic, and imitating relatives' voices to deceive older individuals.