Photographers report that Meta is tagging real photos as ‘Made with AI.’
Meta’s decision to label photos created with AI tools on its social networks started in February, with regular tagging since May across Facebook, Instagram, and Threads apps. However, this approach has faced criticism from users and photographers alike due to instances where the “Made with AI” label was erroneously applied to photos not created using AI tools.
One notable example includes a photo of the Kolkata Knight Riders winning the Indian Premier League Cricket tournament, where the label was incorrectly attached. Importantly, these mislabelings have primarily been observed on Meta’s mobile apps and are not consistently visible on the web platform.
This misapplication of labels highlights ongoing challenges in implementing accurate AI detection and underscores the complexities involved in distinguishing between AI-generated and traditionally captured content. As Meta navigates user feedback and refines its labeling processes, it aims to provide clearer distinctions to maintain transparency and accuracy in how AI-generated content is identified across its platforms.
Many photographers, including former White House photographer Pete Souza, have voiced concerns over their images being incorrectly tagged with the “Made with AI” label by Meta. Their contention is that simply editing a photo with tools, such as Adobe’s cropping tool, should not warrant such a designation.
Pete Souza highlighted his experience in an Instagram post where one of his photos was tagged with the “Made with AI” label. He explained to TechCrunch via email that Adobe recently made changes to its cropping tool, requiring users to “flatten the image” before saving it as a JPEG. Souza suspects that this procedural step triggered Meta’s algorithm to automatically attach the AI label to his photo.
This incident underscores the unintended consequences of automated AI detection systems, which can misidentify photo editing processes as AI-generated content. Many photographers argue that standard photo editing practices, including cropping and basic adjustments, should not result in labels suggesting AI involvement.
As Meta addresses these concerns, it faces the challenge of refining its AI detection algorithms to accurately distinguish between true AI-generated content and photos edited through conventional methods. This effort is crucial to maintaining transparency and avoiding mislabeling issues that affect photographers and users alike on its platforms.
