Several photographers have shared examples over the past few months, with Meta recently labeling a basketball game photo taken by former White House photographer Pete Souza as AI-generated. In another recent example, Meta mistakenly added a tag to an Instagram photo of Kolkata Knight Riders winning the Indian Premier League. Interestingly, like Souza’s photo, the tag only shows up when the image is viewed on a mobile device, not the web.
Suza said he tried to uncheck the tag but failed. He speculated that using Adobe’s cropping tool and flattening the image before saving it as a JPEG image might trigger Meta’s algorithm.
However, Meta also incorrectly labeled real photos as AI when photographers used generative AI tools like Adobe’s Generative Fill to remove the smallest objects, Gigapixel Report. The publication tested this itself using Photoshop’s Generate Fill tool to remove blobs from the image, which Meta then flagged as AI-generated on Instagram. But strangely, Meta did not add the “Made with AI” label when it was released. Gigapixel Upload the file back to Photoshop, then copy and paste it into the black file and save it.
Several photographers expressed displeasure that such minor edits were unfairly labeled as having been produced by artificial intelligence.
“If ‘retouched’ photos are ‘artificial intelligence-made,’ then the term actually means nothing,” photographer Noah Kalina wrote on Threads. “If they really want to protect people, they might as well give every photo an automatic label ‘not a true representation of reality.'”
in a statement edgeMeta spokesperson Kate McLaughlin said the company is aware of the issue and is evaluating its approach “to [its] Labels reflect the amount of artificial intelligence used in the image.
“We rely on other companies to include industry-standard metrics in their tool content, so we are actively working with these companies to improve the process so that our labeling approach meets our intent,” McLaughlin added.
In February of this year, Meta announced that it would begin adding “artificial intelligence-made” tags to photos uploaded on Facebook, Instagram and Threads ahead of this year’s election season. Specifically, the company said it will add the tag to AI-generated photos made using tools from Google, OpenAI, Microsoft, Adobe, Midjourney and Shutterstock.
Meta hasn’t revealed exactly what triggers the “AI-made” label, but all of these companies have (or are working on) adding metadata to image files to indicate the use of AI tools, which Meta identifies as AI-generated A way to take photos. For example, Adobe last year launched a content credentials system that began adding information about the origin of content to metadata.
