TikTok sounds the end of recess. The social network unveiled a new version of its Community Rules on Tuesday. Among its objectives: to strongly regulate the publication of content created by artificial intelligence.
These contents have recently invaded the recommendations of many Internet users. Videos that accumulate millions of views remarkably reproduce the voices of well-known personalities such as Donald Trump, Joe Biden, or fictional characters such as the Dragon Ball hero, to put them in parodic situations -playing Minecraft, in an amusement park, or arguing-. your favorite cartoons.
The platform, threatened with a veto in the United States, has added to its regulations a new section dedicated to “synthetic and manipulated media”, as the US media have warned. the edge. In this section, TikTok declares in particular to prohibit “synthetic media that present the portrait of a private personality”.
In short, it will soon be prohibited to use systems like Midjourney’s image creation AI, or Eleven Labs’ voice playback, to imitate the face or voice of anonymous people on TikTok. These new rules will go into effect on April 21.
An exception for public figures
TikTok, however, makes an exception for “public figures with a significant public role,” such as “a politician, a business leader [ou] a celebrity.” Videos that use artificial intelligence to imitate you may still be posted, except if “the content is used for promotional purposes or violates any other policy,” including hate speech, content of a sexual or harassing nature.
TikTok calls on its users to report deepfakes that do not mention their artificiality. “Synthetic or manipulated media depicting realistic scenes should be prominently displayed,” the Community Guidelines state, which includes recommending the use of a “synthetic,” “fake,” “unreal,” or “modified” label or title for transparency towards users.
Source: BFM TV
