China is trying to prevent deepfakes with new law requiring that videos using AI are prominently marked
- China is making it illegal for fake news to be created with deepfake video and audio.
- "Deepfakes" are video or audio content that have been manipulated using AI to make it look like someone said or did something they have never done.
- The Cyberspace Administration of China (CAC) said that deepfake content can create "political risks" and bring "negative impact to national security and social stability."
- It's unclear exactly how the CAC will enforce the new regulation against deepfakes, as they are notoriously difficult to detect.
- Visit Business Insider's homepage for more stories.
The Cyberspace Administration of China (CAC) announced on Friday that it is making it illegal for fake news to be created with deepfake video and audio, according to Reuters.
"Deepfakes" are video or audio content that have been manipulated using AI to make it look like someone said or did something they have never done.
In its statement, the CAC said "With the adoption of new technologies, such as deepfake, in online video and audio industries, there have been risks in using such content to disrupt social order and violate people's interests, creating political risks and bringing a negative impact to national security and social stability," according to the South China Morning Post reporting on the new regulations.
The CAC's regulations, which go into effect on January 1, 2020, require publishers of deepfake content to disclose that a piece of content is, indeed, a deepfake. It also requires content providers to detect deepfake content themselves, according to the South China Morning Post.
In October, California Governor Gavin Newsom signed two deepfake bills into state law. One of the laws criminalizes publishing deepfake videos involving political candidates 60 days before an election after a deepfake video on Nancy Pelosi went viral. The second law allows anyone to sue someone who publishes a deepfake with their likeness attached to pornagraphic material without their consent.
It's unclear exactly how the CAC will enforce the new regulation against deepfakes, as they are notoriously difficult to detect.
There have been several notable examples that show off a deepfake video's effectiveness at putting words into someone else's mouth, including a deepfake video of Barrack Obama and Mark Zuckerberg making unlikely comments: