TikTok has announced that it will start automatically labeling artificial intelligence-generated content (AIGC) uploaded from other platforms to combat misinformation on its app. This change, aimed at increasing transparency for viewers, will be implemented using technology from the Coalition for Content Provenance and Authenticity (C2PA). By attaching metadata to indicate AI-generated content, TikTok intends to help users distinguish between AI and human-generated content, a critical step given the growing prevalence of AI-generated material on the platform.
This initiative is part of TikTok’s broader effort to address the spread of misinformation, particularly as the U.S. presidential election approaches. Recently, fake AI-generated images of celebrities like Katy Perry and Rihanna at the Met Gala fooled many users, illustrating the challenges posed by AI-generated content. TikTok’s labeling system aims to prevent such instances of misleading content by automatically flagging videos that contain AI elements, enabling users to make more informed decisions about the content they consume.
TikTok’s partnership with C2PA includes using its “content credential” technology, which adds metadata to AI-generated content to indicate its source. While this approach provides a degree of transparency, experts warn that some labels can be easily cropped or removed, suggesting that additional strategies may be needed to maintain credibility. To address these challenges, TikTok plans to produce 12 educational videos in collaboration with Mediawise, a Poynter Institute program, to promote media literacy and help users discern misinformation.
The move to label AI-generated content isn’t unique to TikTok. Other tech companies, such as Google and Meta, have also adopted similar measures to enhance transparency, especially ahead of major political events. TikTok’s engagement with initiatives like the Adobe-led Content Authenticity Initiative further underscores the tech industry’s commitment to promoting truth and transparency in digital media. Ultimately, while these steps represent progress, continuous education and adaptive strategies will be essential to effectively manage the complexities of AI-generated content and its potential for misinformation.