After Meta started tagging photos with a “Made with AI” label in May, photographers complained that the social networking company had been applying labels to authentic images where they had used some essential editing tools.
User Feedback and Tag Modification
Because of user feedback and general confusion about what level of AI is used in a photo, the company is changing the tag to “AI info” across all of Meta’s apps.
Clarifying the Meta AI Tag
Meta said the earlier tag version wasn’t clear enough for users to indicate that the image with the tag was not necessarily created with AI but might have used AI-powered tools in the editing process.
Addressing Misalignment and Context
“Like others across the industry, we’ve found that our labels based on these indicators weren’t always aligned with people’s expectations and didn’t always provide enough context. For example, some content that included minor AI modifications, such as retouching tools, included industry-standard indicators. That was then labeled ‘Made with AI’,” the company said in an updated blog post.
Technology Behind AI Detection
The company is not changing the underlying technology to detect and label the use of AI in photos. Meta still uses information from technical metadata standards such as C2PA and IPTC that include information about the use of AI tools.
Meta Impact on Edited Photos
Suppose photographers use tools like Adobe’s Generative AI Fill to remove objects. Their photos might still be tagged with the new label. However, Meta hopes the new label will help people understand the image. The tag is not always created entirely by AI.
Statement from Meta
AI Info’ can encompass content made and modified with AI, so the hope is that this is more aligned with people’s expectations. At the same time, we work with companies across the industry to improve the process,” Meta spokesperson Kate McLaughlin told TOPCLAPS over email.
Challenges and Limitations
The new tag will not solve the problem of entirely AI-generated photos going undetected. And it won’t tell users how much AI-powered editing has been done on an image.
Setting Guidelines and Fairness
Meta and other social networks must work to set guidelines without being unfair to photographers. Who has not altered their editing workflows? But whose tools are used to touch up photos and have some generative AI element? On the other hand, companies like Adobe should warn photographers. When they use a specific tool, their image might be tagged with a label on different services.
