TikTok to label images and video generated by AI

Sanjana Dhar
Sanjana Dhar May 9, 2024
Updated 2024/05/09 at 3:06 PM

On Thursday, TikTok said that it would start using a technology which will be used to label images and videos generated by artificial intelligence.

The video sharing service announced that it would adopt “Content Credentials”. Content credentials is a digital watermark that is used to denote how the images or videos are edited and created. The technology of Content Credential was spearheaded by Adobe. However it is available to be used by other companies and has already been adopted by certain companies such as ChatGPT creator, OpenAI.

Why is TikTok taking this measure?

With the US elections this fall, researchers have expressed their concern about how content generated by AI could be used as misinformation to interfere with the election procedures. Earlier this year a group of 20 tech companies signed an accord pledging to fight misinformation regarding US elections. TikTok was one of them. Other companies who are also planning to use Content Credentials include YouTube owned by Google and Facebook owned by Meta.

How will Content Credentials work?

For the system to work, the maker of the generative AI tool that is used to make content as well as the platform that is used to distribute the contents generated must agree to use the industry standard.

For example if a person uses OpenAI’s DALL-E tool to generate an image, OpenAI will attach a watermark to the image generated. If the same image is then uploaded on TikTok it will be automatically labeled as AI-generated.

Content generated by artificial intelligence made with tools inside the app has already been labeled as AI generated by TikTok. But with this latest feature, TikTok will now apply a label to AI generated content, made with tools outside the service.

 

For more information please keep reading techinnews

 

 

Share this Article