TikTok regulates posting of AI-powered ‘deepfake’ content and content related to political issues

MADRID, March 21 (Portaltic/EP) –

TikTok has announced a updating their content moderation policies to include new restrictions in the regulation of ‘deepfakes’who will have to notify that it is “false” content, and in the content related to political affairssince the parties will not be able to pay to publish promotional material.

TikTok includes the ‘deepfake’ in its ‘synthetic and manipulated media’ section, a type of content that is based on videos generated by Artificial Intelligence (AI) that mimic very accurately the appearance and voice of a person.

These types of videos are having a great reception on the ByteDance platform, but they pose an added difficulty for users when it comes to recognizing what is real and what is fiction.

In this sense, TikTok has announced news in its content policies with an update, which will come into force on April 23, and which contemplates that content creators point out which posts are ‘deepfake’to let the user know that what they are seeing is not real, as explained on their community standards page.

To do this, the platform has indicated that content creators will be able to use ‘stickers’ or legends in publications of this style that warn that it is AI-generated content and, therefore, be warned that it is “fake”, “not real or “altered“.

The company has introduced it in a new module of its policies titled ‘Synthetic and Manipulated Media’. Thus, according to the platform, TikTok remains open to the “potential for creativity that new AI technologies and other digital technologies can bring with them.” However, it has also indicated that, with this measure, is procured balancing the value of a “synthetic” medium in terms of “expression potential and the risk of harm to individuals”.

Following this thread, TikTok has underlined that will not allow any synthetic media to share the “representation or simulation” of a private person. Specifically, he has specified that in the case of a public person the platform is “more flexible”, but nor will it be allowed to “be abused”, nor that misleading information is disseminated on “political or economic” issues.

The social network emphasizes that these measures also include the prohibition of ‘deepfake’ content that are based on hate speech, sexual exploitation or “severe forms” of harassment.

However, the ‘app’ does allow synthetic media that show a public person in certain contexts. For example, it refers to artistic and educational situations.


The Chinese social network has also regulated specific guidelines on content relating to political affairs. Thus, he has expanded the issues that were previously addressed in his policies on misinformation or false publication of election information.

As he has pointed out, from ByteDance they try to find the “fair balance” between allowing debate and “being a platform for union, not division”. For this reason, TikTok has emphasized that, from now on, the Political parties cannot pay to publish content or political advertising neither too little to collect financing.

This restriction includes from conventional paid ads from application to content creators who are hired to endorse or oppose a candidate in politics.

In addition, nor does it allow misinformation about civic processes or electoral such as, for example, false information about how to vote, about the requirements of the candidates or about the vote counting processes. TikTok has even reported that content that includes any of these false claims will be automatically excluded from the For You feed.

Related articles