In a bid to increase transparency and combat potential misuse of AI-created content, TikTok announced today it will automatically label AI-generated content (AIGC) uploaded from other platforms.
TikTok is partnering with the Coalition for Content Provenance and Authenticity (C2PA) to drive this initiative, leveraging its Content Credentials technology which attaches metadata to content. This will ultimately allow the social platform to identify and label AIGC automatically.
Previously, TikTok labeled AIGC made with its own effects and required creators to disclose externally generated AIGC. Now, this new system expands labeling beyond just TikTok-made AIGC.
The rollout begins with images and videos, and audio content to follow soon. TikTok plans to attach Content Credentials to its own AIGC in the coming months.
The metadata will reveal details about the content’s creation process and remain even after download. Other platforms adopting Content Credentials will then be able to automatically label the content as AIGC.
To enable users to navigate AIGC and fight misinformation, the social platform is launching new media literacy resources developed with experts like Mediawise and WITNESS. These resources aim to educate users on identifying and critically evaluating AIGC content.
“AI-generated content is an incredible creative outlet, but transparency for viewers is critical,” said Adam Presser, TikTok’s Head of Operations and Trust & Safety. “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.”
This initiative makes TikTok the first video-sharing platform to implement Content Credentials. While Meta previously announced plans to utilize C2PA’s solution, TikTok is ahead in actively labeling AIGC from various sources.
The social plafrom is working to combat deceptive AIGC, especially in elections. The platform’s policies already prohibit misleading AI-generated content, regardless of labeling. It aims to continue investing in proactive detection models, expert consultations, and industry collaboration to address this evolving challenge.
Author
