After Meta began labeling photos as “created with AI” in May, photographers complained that the social networking company was labeling actual photos that used basic editing tools.
Due to user feedback and general confusion about what level of AI is used in the photos, the company is changing the tag to “AI Information” across all Meta apps.
Mehta said previous versions of the tag weren't clear enough for users to indicate that a tagged image wasn't necessarily created by AI, but that AI-powered tools might have been used in the editing process.
“Like others across the industry, we found that our labels based on these metrics didn't always align with people's expectations and didn't provide enough context. For example, some content that included minor AI-based changes, such as retouching tools, included industry-standard metrics and was labeled 'Made with AI,'” the company said in an updated blog post.
Image credit: Meta
The company has no plans to change the underlying technology that detects and labels the use of AI in photos: Meta will continue to use information from technical metadata standards such as C2PA and IPTC, which include information about the use of AI tools.
That means if a photographer uses a tool like Adobe's Generative AI Fill to remove an object, the photo could still be given a new label, but Meta hopes the new label will help people understand that a tagged image isn't necessarily created entirely by AI.
“AI Info will include content created and/or modified with AI, so we hope this will align with what people have come to expect. At the same time, we're working with companies across the industry to improve our processes,” Meta spokesperson Kate McLaughlin told TechCrunch in an email.
The new tags don't solve the problem of not detecting photos that are entirely AI-generated, nor do they tell users how much AI editing has been done to an image.
Meta and other social networks will need to set guidelines to ensure they aren’t unfair to photographers who haven’t changed their editing workflow but the tools they used to touch up their photos include generative AI elements, while companies like Adobe will need to warn photographers that using certain tools could result in their images being labelled on other services.