As AI continues to get more terrifyingly realistic, platforms are working to be a little bit more transparent over what is and isn’t AI-generated. Especially as we get closer to election season.
So in an effort to combat misinformation, Snapchat has become the latest platform to introduce AI watermarks.
What are Snapchat AI watermarks?
Visually, these watermarks appear as a little Snapchat ghost with sparkles next to it. They’re automatically generated when people use Snap’s generative AI features like Dream and AI Snaps. They also incorporate metadata like the time and location of creation.
“The purpose of these watermarks is to provide transparency that the image was created with generative AI and is not real or based on real events, even if it is a realistic style,” Snap wrote in a post.
How do watermarks work?
These watermarks work slightly differently depending on which AI tool Snapchatters use. For instance, if a Snapchatter shares an AI-generated Dreams image with a friend, the recipient will see a context card about how the image was AI-generated.
But for other AI tools, like Snapchat’s AI-powered “extend” tool, the watermark will just be the little ghost.
When making these images or videos, the system is able to generate unique watermarks while incorporating metadata — like the time and location of creation.
This also makes the watermark difficult to tamper with or alter. Furthermore, Snap added that any attempt to tamper with or remove these watermarks would be a violation of its terms of service.
Further Reading:
