Meta Will Soon Label AI-Generated Images Posted on Facebook, Instagram and Threads

Meta announced today that it will soon start labeling AI-generated images on Facebook, Instagram, and Threads. The company said that it’s also working with partners on new standards for identifying AI-generated content across the Internet.

With tools like OpenAI’s DALL·E and Microsoft Designer, it’s now easier than ever to create AI-generated images. And just last week, X/Twitter and Microsoft had to intervene after porn deepfakes of Taylor Swift started circulating online. As the upcoming US presidential election will likely increase the circulation of AI-generated images and videos, Meta wants to do its part to reduce the risks of disinformation.

On its own platforms, Meta already applies “Imagined with AI” labels to photorealistic images created with the company’s Meta AI feature. However, doing something for all AI-generated images will require some cooperation with industry partners.

“We’re building industry-leading tools that can identify invisible markers at scale – specifically, the “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 for adding metadata to images created by their tools,” explained Nick Clegg, President of Global Affairs at Meta.

Meta plans to start applying labels for AI-generated in all languages supported by Facebook, Instagram, and Threads in the coming months. For AI-generated video or audio, however, companies working on generative AI tools have just started to include invisible markers in their tools. Clegg admitted that “we can’t yet detect those signals and label this content from other companies.” For the time being, Meta will soon start asking users posting AI-generated video and audio to manually disclose that their content was created with AI.

“We’ll require people to use this disclosure and label tool when they post organic content with a photorealistic video or realistic-sounding audio that was digitally created or altered, and we may apply penalties if they fail to do so,” the exec explained. “If we determine that digitally created or altered image, video or audio content creates a particularly high risk of materially deceiving the public on a matter of importance, we may add a more prominent label if appropriate, so people have more information and context.”

Google recently adopted a similar approach regarding AI-generated videos on YouTube. Content creators posting videos that have been created or modified with AI will soon be required to disclose it, and YouTube will then add an “altered or synthetic content” label to these videos. Google also plans to penalize content creators who fail to disclose their use of AI tools by demonetizing or removing their videos.