It’s becoming increasingly difficult to determine what’s AI-generated and what’s not. Meta wants to make it easier to spot AI content with just a quick glance.
Meta to label AI-generated content ahead of upcoming elections | Firstpost America
On Tuesday, Meta announced plans to add “visible markup” and metadata indicators to AI-generated content appearing on Facebook, Instagram and Threads. That includes labeling AI-generated images from “Google, OpenAI, Microsoft, Adobe, Midjourney and Shutterstock.” For now, though, those labels will only appear on images; videos and audio present a bigger, trickier problem.
“While companies are starting to incorporate signals into their image generators, they have not yet begun to incorporate them into AI tools that generate audio and video at the same scale. As a result, we cannot yet detect these signals and label this content from other companies,” the company wrote in the blog post.
To address this issue, Meta wants creators to indicate that audio and video were created by AI. If creators don't, they could face "penalties."
