© solenfeyissa/Pixabay
Today, some AI models are so powerful that they are able to generate ultra-realistic, and very misleading, images. And while online platforms generally do not want to prohibit the use of images generated with artificial intelligence, they try to fight against abuse and be transparent.
As for TikTok, for example, it already notifies the user, via a label, when content that has been generated with its own tools appears on the screen. And soon, this system, which allows users to be informed, will be extended to images generated with AI on other online tools.
To identify images that were created with generative artificial intelligence, TikTok will rely on Content Credentials technology from the Coalition for Content Provenance and Authenticity (C2PA). This technology allows image generators to include a hidden signature in the content they produce, which then allows platforms like TikTok to instantly identify AI-generated images. For example, if a person generates an image on DALL-E 3, OpenAI's image creation model, and then posts that image to TikTok, the social network will be able to identify that content as a product of the 'Generative AI.
Limit risks, without harming creativity
“Content generated by the&# 8217;AI is an incredible creative outlet, but transparency for viewers is key”, said Adam Presser, head of operations, trust and safety at TikTok, according to TechCrunch. “By partnering with peers to tag content across platforms, we make it easy for creators to responsibly explore AI-generated content, while continuing to deter harmful or misleading AIGC (editor's note, AI-generated content) which are prohibited on TikTok.”
In addition, in addition to labeling AI-generated content that is published on its platform, TikTok will also apply Content Credentials technology to images that are created with its AI filters. So, if a person produces content with these TikTok filters, this content can be tagged by other social networks.
As a reminder, Meta, which includes Facebook, Instagram and WhatsApp, is also committed to using Content Credentials. “We are developing cutting-edge tools capable of identifying invisible markers at large scale… in particular “AI-generated” information in the C2PA and IPTC technical standards – so we can label images from Google, OpenAI, Microsoft, Adobe, Midjourney and Shutterstock. as they implement their plans to add metadata to the images created by their tools”, Mark Zuckerberg's company announced in February.< /p>
- TikTok will start labeling images generated with AI, even when they were created on other tools , like DALL-E 3
- To identify AI-generated images, TikTok relies on Content Credentials from the Coalition for Content Provenance and Authenticity (C2PA)
- This technology allows AI models to place markers on the content they produce, so that the content is easily identifiable as a product of artificial intelligence
- The Meta group is also committed to offering similar functionality on its platforms
📍 To not miss any news from Presse-citron, follow us on Google News and WhatsApp.
21.1 M reviews
[ ]