TikTok will start detecting AI-generated content

TikTok has announced that it has started automatically tagging user-generated content with artificial intelligence, even if it comes from external applications. To do this, the video network has partnered with the Coalition for Content Provenance and Authenticity (C2PA), which includes Adobe and Microsoft among its founders.

TikTok DOWNLOAD

Although TikTok offers access to a wide variety of AI-powered effects and filters, all videos that use them are already tagged by the platform. Additionally, TikTok requires creators to clearly indicate when they are using realistic CGIA (AI-generated content). To facilitate all of this, the platform built a tool that, according to its own reports, has already been used by over 37 million people.

In a post published by the company, they clarify that “while most people want to enjoy AI-generated content responsibly, there are malicious actors who can use AI to intentionally deceive others.”

How can TikTok detect images and videos created by AI? In this case, it is thanks to Content Credentials. These are a type of metadata that serve to verify the authenticity and origin of digital content. In summary, they are the implementation of the C2PA guidelines.

In this way, TikTok can “read” this “digital signature” in the images and videos that are going to be uploaded to the platform. In the near future, it will also do the same with audios. This has allowed TikTok to proclaim itself as “the first video platform to implement Content Credentials”.

TikTok also warns in its post that tags could cause confusion if viewers don’t have context of what they mean. From the platform, they announce that they are creating awareness campaigns to help the community “think critically about CGIA and misinformation.”

TikTok DOWNLOAD