Machine learning-based deepfake videos have skyrocketed in popularity in recent years. From official use in The Mandalorian to parody videos, the tech is constantly being improved and used in a variety of ways.
Deepfake technology has been criticised for its ability to blur the lines between fact and fiction. For example, anyone can make a celebrity say or do whatever they want. Additionally, the tech has been used to create explicit videos of famous people without their consent.
China starts to take on deepfake tech
Reported by Rest of World, China is starting to tackle the issue of deepfakes. The crackdown comes alongside new regulations surrounding algorithms used by big tech companies in the region. For example, the region is trying to limit algorithmic influences on user behaviour.
However, the regulations also came with a new that will attempt to restrict deepfakes for the first time. As part of the new rules, social media companies are “forbidden” from pushing users towards manipulated media.
The regulations claim that social platforms — Baidu, TikTok, etc cetera — will not be allowed to spread “synthetic content”. This is a large net term that incudes purposely created false information and, of course, deepfake videos.
These regulations don't limit “synthetic content” from existing on platforms. However, the content cannot be recommended by algorithms such as Tiktok’s “For You” scroll. This means that if you want to watch deepfakes, you'll have to search for it.
Is this harming free speech?
The blanket phrase of “synthetic content” can possibly have a wide-reaching impact on non-deepfake content. For example, does “false information” include satirical content, fake news designed for comedy?
China has a long history of censorship, and these regulations could result in even more censorship. While it’s good that dangerous deepfakes will now be limited, the media that gets caught in the crossfire might be needed.