TikTok limits the use of tools to generate images using AI

AS the United States ponders banning the Chinese application, TikTok is betting on bolstered community rules to give itself a new, more serious image. As part of this, the social network is taking a stand on the use of AI tools.

"We want everyone to feel confident that they can access information that is reliable, discover content that is original, and engage with people who are authentic. This is foundational to building a community of trust and accountability both on TikTok and off-platform," says the Chinese social media site.

From ChatGPT to DALL-E, there's no shortage of tools to generate content and images using artificial intelligence.

And TikTok is only too aware of this, and of the fact that its users won't fail to use them.

While bans are clamping down on the use of TikTok in Canada, the United Kingdom, New Zealand and Europe, the Chinese giant seems to want to work on its image in order to appear more reliable and inspire less reticence on the global stage: "In a global community, it is natural for people to have different opinions, but we seek to operate on a shared set of facts and reality.

We do not allow inaccurate, misleading, or false content that may cause significant harm to individuals or society, regardless of intent," read TikTok's updated community guidelines.

As part of this, TikTok has decided to regulate the use of artificial intelligence by prohibiting the use of AI-generated images reproducing the likeness of any private figure.

For public figures, TikTok's policy is more nuanced. If reproductions are allowed, they must not be used for political or financial gain. In other words, users cannot use the AI-generated image of a well-known personality for profit.

"While we provide more latitude for public figures, we do not want them to be the subject of abuse," says TikTok. "We do not allow synthetic media of public figures if the content is used for endorsements or violates any other policy. This includes prohibitions on hate speech, sexual exploitation, and serious forms of harassment."

This kind of "synthetic" content generated by artificial intelligence will also have to be explicitly labelled as such. Creators will have to add "synthetic" or "fake," "not real" or "altered" in their caption: "AI can make it more difficult to distinguish between fact and fiction, carrying both societal and individual risks," says TikTok.

Previously, deepfakes of Tom Cruise caused a viral buzz on TikTok, leading to confusion among internet users. At the time, nothing indicated that the content was deepfake.

Since then, the account @deeptomcruise has added "parody" to its bio, but the captions do not yet contain the mandatory terms listed in TikTok's guidelines. The new rules will take effect on April 21, 2023.