We’re in the middle of an AI era right now, and while Apple continues to push for AI integration across hardware and services (including the latest M4 chip on the iPad Pro), the world of generative AI is a metaphorical minefield.
From AI models being trained on existing art to the potential for deceptive depictions of real-world figures, there’s a lot to unpack. And, while TikTok has been in the news for potentially being banned, the app is taking a big step to clear things up when it comes to AI-generated content.
The company announced on Thursday that it would look to “advance AI transparency and literacy”. As per the press release, TikTok already labels AI-generated content (AIGC) made with its own AI effects, and “realistic AIGC” has had to be labelled by creators for over a year.
Now, TikTok will automatically label AIGC through its Content Credentials, but it’s not just the content that’s been made within TikTok that’ll be flagged. These Content Credentials are a technology from the Coalition for Content Provenance and Authenticity (C2PA) and help identify video and images with AI metadata.
Industry-wide AI transparency
TikTok’s press release explains it’s already begun labelling AI images and videos, and will come to audio-only content soon.
“Over the coming months, we’ll also start attaching Content Credentials to TikTok content, which will remain on content when downloaded,” the release explains.
“That means that anyone will be able to use C2PA’s Verify tool to help identify AIGC that was made on TikTok and even learn when, where and how the content was made or edited.”
This will also roll out to other social networks as they adopt the labelling process, but it should help flag when images, videos, and eventually audio are constructed using AI.
