Connect with us

Tech

Meta Switches AI Labels From ‘Made with AI’ to ‘AI Info’ Tag

Published

on

Meta Switches AI Labels From ‘Made with AI’ to ‘AI Info’ Tag

Meta’s making a slight tweak to its “Made with AI” labels, in order to clarify confusion around what, exactly, these indicators mean when displayed in-stream.

Apparently, some Facebook and IG users have been less than pleased that their uploads have been tagged as “Made with AI”, when they’ve only used digital tools for slight re-touching. As such, Meta’s now revised the wording of the tags to a less definitive “AI Info” instead.

As you can see in these examples, now, when Meta’s system detects that AI tools have been used to create an image, or when users indicate that they’ve used AI tools in the composition, the tag will include an “AI Info” note, which, when tapped, will provide an overview of how AI could have been used in creating the image.

As per Meta:

We’ve found that our labels based on [industry standard AI] indicators weren’t always aligned with people’s expectations and didn’t always provide enough context. For example, some content that included minor modifications using AI, such as retouching tools, included industry standard indicators that were then labeled “Made with AI.” While we work with companies across the industry to improve the process so our labeling approach better matches our intent, we’re updating the “Made with AI” label to “AI info” across our apps, which people can click for more information.

So rather than the blanket flag that an image has been AI generated, Meta’s revised wording will soften the impact of the tags, so that users who’ve only applied some retouching won’t get criticized for posting AI content.

Which is especially important for artists and photographers, whose work has been tagged as AI, when it isn’t. That’s led to accusations and arguments in the comments, forcing people to justify themselves for their creations.

It’s another indicator of the evolving AI landscape, and the various challenges it will prompt in regards to how we determine what’s real, and what isn’t, and what we trust in terms of creator skill versus machine generation.

And indeed, whether that’s even important moving forward. Right now, there’s a real pushback to AI art, which is created from text prompts, but eventually, these generations are likely to become so commonplace that the divide between what’s real and what’s not may not even be relevant.

But right now at least, there’s a need to better define the actual contribution of AI tools in visual creation.

I guess, the counter to that is that users who want to present AI creations as their own work will now use this as a means to play down how they’ve used AI in their creation process, but it does make sense for Meta to provide a broader qualification of the process to appease human creators.

Continue Reading