Occurred: October 2023
A realistic deepfake video of popstar Taylor Swift fluently speaking Mandarin Chinese sparked discussion in China about the ethics of using artificial intelligence to develop digital content.
Developed by Chinese American AI video creation company HeyGen, the video shows Swift flaunting her Mandarin in what looks like a talk show. But the clip appeared largely designed to promote itself than say anything about Taylor Swift, who does not speak Mandarin.
While some Chinese citizens and commentators praised the quality of the video, others condemned it for its apparent lack of consent from Swift and the potential loss of privacy and damage to an individual or organisation's name, image and reputation caused by non-consensual synthetic media.
On its website, HeyGen states it is a member of the Content Authenticity Initiative, which promotes 'authentic storytelling' and 'promotes transparency' in the use of AI.
Country: China; USA
Purpose: Promote developer
Technology: Deepfake - audio, video; Generative adversarial network (GAN); Neural network; Deep learning; Machine learning
Issue: Mis/disinformation; Privacy
News, commentary, analysis
Published: November 2023