advertisement
Facebook
X
LinkedIn
WhatsApp
Reddit

TikTok has a new plan for AI content

  • TikTok will now be labelling content that is generated by AI and then posted on the platform.
  • It will be leveraging the “Content Credentials” watermark, which is used by members of a coalition seeking to label all AI content online.
  • This coalition includes the likes of OpenAI, Microsoft and Google, as well as other major AI players.

As surprisingly authentic text, images, and video generated by AI are now flooding the internet at an unforeseen and unstoppable pace, certain companies are looking to make these pieces of content stand out from the rest, including social media TikTok.

The company said last week that it would begin tagging AI-generated content making rounds on its video-sharing platform with a digital watermark called “Content Credentials”. This latest watermark would be applied to content generated outside of TikTok and then reposted on the app.

TikTok already labels content generated with AI software inside the app, such as filters than use AI to generate different images of your face, for example.

“We also have policies that prohibit realistic AI that is not labeled, so if realistic AI (generated contents) appears on the platform, then we will remove it as violating our community guidelines,” Adam Presser, head of operations and trust and safety at TikTok, told Reuters.

According to Reuters, TikTok’s watermark technology for AI generated content actually stems from tech created by Adobe, Microsoft and other members of the Coalition for Content Provenance and Authenticity.

The coalition represents a group of major corporations that are working to highlight when a piece of content is made by AI and thus help slow the spread of fake news or misinformation online. The technology works between companies, for example, if OpenAI’s Dall-E is used to generate an image and then posted to TikTok, the platform will automatically apply a label to it.

This is because Dall-E automatically adds an invisible watermark in the data of the image generated by Dalle-E, which marks that it is AI-generated, and OpenAI is part of the coalition, alongside TikTok. A similar system is being used by Google to mark content generated by its own image generators.

Google itself is a member of the coalition, which includes most big players in generative AI, except for Stability.AI, the creators of Stable Diffusion, one of the world’s most popular and easily accessible image generators.

TikTok is currently on its way to the hangman in the United States, as the platform has until next year to find an American buyer or be forced to pull out of the country, becoming unavailable to all Americans unless they have a VPN.

However, the latest from ByteDance, the owner of TikTok, is that it has no intention of selling the platform, and thus will most likely depart the country of its own will unless the United States government runs back its TikTok ban bill.

[Image – Photo by Nik on Unsplash]

advertisement

About Author

advertisement

Related News

advertisement