TikTok automatically labels AI content from platforms like DALL·E 3

TikTok begins automatically labeling AI-generated content from other platforms. Any video/photo posted to TikTok, if using a service like OpenAI’s DALL·E 3, will automatically have an “AI-generated” label to notify viewers.

This video platform uses Content Credentials, a technology from the Content Provenance and Authenticity Alliance (C2PA) – an organization co-founded by Microsoft and Adobe. Content Credentials adds detailed information to content, helping TikTok automatically identify and label AI content.

Therefore, TikTok will automatically label uploaded AI content with Content Credentials. This change begins Thursday and will apply to all users globally in the coming weeks.

While TikTok already labels content that uses its own AI effects, it will now label content from other platforms with Content Credentials, such as OpenAI’s DALL·E 3 and Bing Image Creator by Microsoft. Microsoft, Adobe and OpenAI have joined Content Credentials, and Google has also pledged its support.

While TikTok requires content creators to disclose when posting content created or edited with AI, the company says this new change is an additional way to ensure AI content is labeled, while also reducing Pressure on creators.

In the coming months, TikTok will also begin attaching Content Credentials to AI content created on their platform using TikTok AI effects. Details in Content Credentials will include where and how to create or edit AI content, and will be preserved when downloaded. Other platforms that adopt Content Credentials can also automatically label AI content.

In short, TikTok is committed to labeling AI content on its service and also wants to ensure AI content created on TikTok is labeled correctly when posted on other platforms.

“AI-generated content is a great creative output, but transparency for viewers is critical,” Adam Presser, Director of Operations and Trust & Safety at TikTok, said in a press release will. “By working with partners to label content across platforms, we make it easier for creators to discover AI content responsibly, while continuing to prevent AI content that is harmful or misleading is banned on TikTok.”

TikTok is proud to be the first video sharing platform to implement Content Credentials. Notably, Meta announced in February that it plans to build on C2PA’s solution to add provenance to content.

As part of Thursday’s announcement, TikTok said it is committed to combating the use of deceptive AI in elections and that its policies prohibit AI-generated content that is misleading or harmful — regardless of labeled or not.