TikTok Introducing Automatic Labeling of AI-Generated Content

TikTok has announced a new feature that automatically identifies and labels AI-generated content made from external platforms. Beginning Thursday, the popular social video platform will attach an “AI-generated” label to content created using services like OpenAI’s DALL·E 3, alerting viewers to its origin.

This initiative is made possible by integrating Content Credentials, a technology developed by the Coalition for Content Provenance and Authenticity (C2PA), a collaboration between Microsoft and Adobe. Content Credentials embed specific metadata into content, enabling TikTok to swiftly identify and categorize AI-generated content.

As part of its ongoing commitment to transparency, TikTok will now automatically apply the “AI-generated” label to such content uploaded to the platform, complete with Content Credentials. This change will be implemented globally starting this Thursday and will gradually extend to all users over the course of the following weeks.

Even though TikTok already marks content produced with its own AI effects, it will extend this labeling to content generated on other platforms utilizing Content Credentials, such as OpenAI’s DALL·E 3 and Microsoft’s Bing Image Creator. While Microsoft, Adobe, and OpenAI are already aligned with Content Credentials, Google has also committed to supporting Content Credentials.

Tiktok AI generated content
Image Credits: TikTok

Although TikTok already asks creators to disclose AI-enhanced content, the company perceives the new change as an extra measure to ensure proper labeling of AI-generated content while reducing the burden on creators.

In the upcoming months, TikTok will also begin integrating Content Credentials into AI-generated content produced on the platform using TikTok AI effects. This metadata will provide information on the origin and editing process of the AI-generated content and will be attacked even when downloaded. Other platforms adopting Content Credentials will be equipped to label such content as AI-generated automatically.

Read More: OpenAI Introduces Image Detection Tool To Detect DALL-E 3 Creations

Therefore, while TikTok is committed to labeling AI content within its own ecosystem, it is also aiming to facilitate accurate labeling of AI content originating from TikTok when shared on other platforms.

In a press release, Adam Presser, head of Operations and Trust & Safety at TikTok, said, “AI-generated content is an incredible creative outlet, but transparency for viewers is critical. By partnering with peers to label content across platforms, we’re making it easy for creators to responsibly explore AI-generated content, while continuing to deter the harmful or misleading AIGC that is prohibited on TikTok.”

TikTok proudly asserts its status as the first video-sharing platform to integrate Content Credentials. Notably, Meta revealed in February its intentions to expand upon the C2PA’s framework for embedding content origin.

In its latest announcement on Thursday, TikTok emphasized its dedication to combating the misuse of deceptive AI, particularly in elections. The platform underscored its strict policies, which unquestionably forbid the circulation of misleading AI-generated content, irrespective of whether it bears labeling.