China cracks down on fake news and deepfakes with new laws

Brittany A. Roston - Dec 1, 2019, 11:24 am CST
2
China cracks down on fake news and deepfakes with new laws

China is cracking down on the creation of fake news and ‘deepfakes’ with new laws that will make publishing this content without an advisory a criminal offense. The move comes amid growing concerns over the effects AI-generated fake content can have on spreading false information and propaganda across social platforms. The move follows a similar bill recently passed in California.

READ: Gaming addiction in China leads to heavy curfew for minors

Deepfakes is a term that typically refers to videos modified to feature the face of one individual placed realistically on the body of another individual.

Though the results aren’t always realistic, many deepfake videos are very convincing, particularly if shared at lower resolutions, making it possible to trick people into thinking that a public figure said something they never actually expressed.

The technology introduces a number of concerns; we’ve already seen examples of people like former President Barack Obama being presented as saying things he never said via the magic of artificial intelligence. Likewise, this technology has been maliciously used to generate adult content featuring the likeness of people who never consented to the work.

A number of researchers with various institutions have worked on developing new technologies that can analyze and detect when a video or image has been altered using software. These systems, however, are largely restricted to academic and commercial fields; they haven’t trickled down to the consumer level in any significant way, leaving many people without any simple way to tell whether a particular news article or piece of content is authentic.

China is addressing the matter by making it a criminal offense to publish and distribute deepfakes and fake news without prominently marking it as inauthentic, according to Reuters. Chinese officials cite concerns that this type of generated content may ‘endanger national security, disrupt social stability, disrupt social order and infringe upon the legitimate rights and interests of others.’ The new laws will go into effect on January 1, 2020.


Must Read Bits & Bytes