Meta is adding Made by AI label to real photos, photographers ...
In March 2023, viral images of Pope Francis wearing a white puffer jacket and shades stirred conversations worldwide. These images turned out to be deepfakes, leaving many astonished at the realistic AI generation. Fast forward a year, the reality of AI-generated content is well-known, prompting companies to combat this issue. One such company, Meta, recently began labeling AI-generated images. However, this endeavor has faced challenges, with photographers expressing discontent over authentic images being mistakenly labelled.
Challenges Faced by Meta
According to a report by TechCrunch, Meta, the parent company of Facebook and Instagram, has received criticism from users and photographers for incorrectly tagging real photos as "Made with AI." The introduction of Meta's new AI detection system has led to widespread frustration, as many individuals argue that their human-crafted or minimally edited images are being inaccurately identified.
Meta announced its plan in February to label photos created using AI tools on its social platforms. The goal was to assist users in distinguishing between AI-generated and authentic content. Since May, Meta has routinely applied the "Made with AI" label to certain images on Facebook, Instagram, and Threads. Despite initial intentions, this initiative has come under scrutiny for its mislabeling of images, prompting backlash from photographers and users.
Photographers' Experiences
Pete Souza, a former White House photographer, shared his firsthand encounter with Meta's labeling issue. He mentioned that one of his photos, a digital scan of a film image from the 1984 NBA Finals, received the new label despite being a non-AI creation. Souza speculated that the process of "flattening the image" before saving it as a JPEG in Adobe's software might have triggered Meta's algorithm to assign the AI tag. Expressing his frustration, Souza noted the forced inclusion of the tag even after attempting to remove it.
Reports indicate that other users have faced similar challenges. A cosplay artist had their project image incorrectly labeled as AI-generated, despite minimal editing for color correction. Another artist, specializing in pixel-based digital art, also reported mislabeling of their work.
Responding to the feedback, Meta is reassessing its labeling approach to ensure accurate reflection of AI usage in images. A Meta spokesperson emphasized, "Our intent has always been to help people identify AI-generated content. We are incorporating recent feedback and continuously evaluating our strategy."
