Tiktok’s new directives add subtle changes to live creators, AI content, and even more

Tiktok is preparing an update of its community guidelines, which dictate the rules of participation in the social video platform as well as the standards that the company uses to determine the videos that feed it for you. Although the update largely rewrites the original text for the simplicity of simplicity, there are some elements that jump into the new revision – in particular how Tiktok prioritizes its market, the depth of its experience according to the experience and a small change to the type of content generated by Ai -Auté.

Today, social media companies must comply with a range of regulations on a global scale, such as the UK online security law, the EU digital service law (DSA) and the American law on care, which, in part, has led to political revisions on certain platforms, as with the counterpart of yesterday Bluesky update.

The changes of Tiktok, which are online on September 13, 2025, are not as substantial, because it seems that the company often rewrites text for more clarity.

However, a section that has seen several additions in the updated directives covers the Rules for the creators of Live Tiktok.

The company warns the creators that they are responsible for everything that is happening during their live session, even if it involves third -party tools, such as real -time or text translation tools to read the comments of viewers. Tiktok advises creators to monitor these tools to ensure that they do not violate the rules thanks to these third -party services.

Another notable addition to this section presents new directives for commercial content.

Tiktok stresses that the commercial content must be disclosed. He too declare directly that it will reduce the visibility of the content that orders users to “buy products outside the platform on the markets where Tiktok Shop is available”.

The company also says that it is Personalization of user search results.

Although the previous version of the directives said that Tiktok provided “research suggestions” that were relevant to the user, new guidelines say that “research results and recommendations may seem different for everyone”. The guidelines explain that Tiktok uses information like your past research and what you looked at to make the search results more relevant.

Updated rules reveal that comments are personalized too.

Tiktok says that the comments are sorted according to signals such as answers, tastes and past reports. Again, this means that the comments section will be different from the user to the user.

THE AI content section has not changed spectacularly, although the language here has become less verbose in part during the description of the Deepfake content type is not authorized. Tiktok previously prohibited the content that “shares or shows false sources or crisis events, or falsely shows public figures in certain contexts,” he said. “This understands being intimidated, making an approval or being approved.”

This has now been replaced by a language which says that Tiktok does not allow the content “which is misleading on questions of public importance or harmful for individuals”.

It is interesting to note that the language referring to AI approvals has been deleted. (Maybe Tiktok opens the way to endorsements approved by celebrities and generated by celebrities?)

In some cases, the language of guidelines has been simplified, as in the case of For you, the eligibility standards section (FYF)who no longer has A long list of what is considered to be ineligible fyf content. Instead, the details on the ineligible content are distributed in the many different sections of the updated community directives, which makes it less useful to refer because it is no longer in the same place.

It should also be noted that Tiktok has changed the language explaining the reason why it is involved in the moderation of the content.

Before, the company said that the process had maintained the “safe, trustworthy and dynamic” platform. Now he describes the moderation of the content as caregiver to be a “place of security and creation for everyone”.

It seems that “trustworthy” has obtained the ax. UH-OH.