Research

China wants to ban “dipfaces” - fake videos created using neural networks

The web is increasingly distributed videos, allegedlyrecorded with the participation of a famous person - but she is not known. Fake videos demonstrate a person in a situation that never existed, and sometimes it seems impossible to distinguish a fake. Such videos are created with the participation of the GAN neural networks (a generative-adversary network) from almost nothing: just a photo downloaded from the Internet. China decided to put an end to the spread of such “deepfakes”.

China bans fake videos

China plans to legally ban“Dipfeyki” - and this is the first country that took seriously the manipulation of artificial intelligence from the video and sees them as a threat. On Saturday, the highest legislative body of China proposed to declare illegal “distortion of the image of another person or imitation of voice using technology,” according to a material published in the state newspaper China Daily.

Newspaper quotes senior lawmaker ShenChunyao, who says that “the misuse of artificial intelligence technology not only harms human rights in the image, but also damages national security and public interests.”

"Deep fakes," or "dipfeyki" becomemore and more sophisticated - instead of simply replacing a person’s face, coders have learned to replace the whole body. This means that anyone with technical know-how can transform anyone into a digital puppet.

If you do not understand how it can be dangerous,In 2018, a fake video appeared with a speech by former US President Barack Obama, where he warned that "we are entering an era in which our enemies can make anyone say anything at any time." The video itself with this statement is a fake, this is an incident.

Only here, China regarded this warning as quite real. Does this technology seem dangerous to you? Tell us in our chat in Telegram.