Of the on January 10, 2023,a new law framing the “synthesis technologies” will come into force in China. This expression refers to the deepfake or hyperfaking, a technique for manipulating audio and video content using artificial intelligence. This technology can be used to propaganda and disinformation purposes. China wants to protect itself against uses that run counter to its interests. China Cyberspace Administration, authority in charge of Internet control, is at the origin of this regulation.
The law includes several major proposals:
Users must provide consent if their image is to be used in deep synthesis technology.
Deep digest services cannot use technology to spread fake news.
Content that violates applicable laws is prohibited, as is content that endangers national interests and security, damages the national image, or disrupts the economy.
Little by little, Chinese power is making its nest
One regulation after another, the Chinese government is tightening its grip around the technology sector. Tech giants, like Tencent and Alibaba, have seen their business suffer from Beijing’s repression in the digital domain. Since 2020, several laws, including a regulation antitrust, or on the use of personal data, aim to regulate the activities of Chinese tech players. At the same time, China also wants to anticipate the expansion of emerging technologies, by regulating the use of private facial recognition and deepfakes.
The Chinese authorities want to take advantage of technological tools by putting them at the service of the Party’s vision. Political power thus hopes to build a stable and prosperous society, in accordance with the Chinese social contract: a constant improvement in the standard of living against the absence of questioning of the legitimacy of the institutions. Above all, the regime wishes prevent any dispute, in particular virtual, by full control of the Internet.
However, Beijing wants to keep its place in the global technological race and does not give up the use of hyper special effects. Companies will therefore have to comply with the provisions of the text, and will be subject to a security audit.
The difficult regulation of deepfakes
Two researchers judge the future law incomplete. Among other things, the procedure obtaining consent of the individual whose image is exploited is not explained and the labeling of deepfakes, made compulsory, can easily be modified by a new encoding.
The coming months will be rich in lessons for the Xi Jinping regime, as well as for other powers, such as the United States and the European Union, where the regulation of deepfake is still incomplete.
We want to thank the author of this write-up for this remarkable web content
