Meta’s recent initiative to label photos created with AI tools on its platforms has ignited backlash among photographers. Since May, photos on Facebook, Instagram, and Threads have been tagged with the “Made with AI” label, indicating they were generated or heavily edited using AI technologies. However, many photographers have reported that their authentic, non-AI photos are being incorrectly tagged, causing confusion and frustration.
Photographers Protest Incorrect ‘Made with AI’ Label
Numerous photographers have voiced their concerns about Meta’s automated labeling system misclassifying their images. For instance, a photo of the Kolkata Knight Riders winning the Indian Premier League was erroneously tagged as “Made with AI,” despite no AI tools being used in its creation. The mislabeling issue is prevalent on Meta’s mobile apps, though it doesn’t appear on the web versions.
Former White House photographer Pete Souza highlighted the problem when one of his images was tagged with the new label. He suggested that the label might have been triggered by the use of Adobe’s cropping tool, which involves flattening the image before saving it as a JPEG. Souza criticized Meta’s labeling process, pointing out that even when he attempted to remove the “Made with AI” tag, it persisted.
Meta’s Response to the Controversy
Meta has not officially responded to specific incidents of mislabeling, leaving many photographers without answers. In a February blog post, Meta explained that their system uses image metadata to identify and label AI-generated content. They rely on standards like C2PA and IPTC to detect AI involvement from tools by companies like Google, OpenAI, and Adobe.
Recent reports suggest that the label is applied when tools like Adobe’s Generative AI Fill are used to alter images. While some photographers agree that any AI usage should be disclosed, others argue that minor edits shouldn’t warrant an AI label. The ongoing debate highlights the need for clearer guidelines and more accurate detection methods to ensure genuine photos are not mislabeled, maintaining trust and transparency on Meta’s platforms.
Read more: Meta Introduces AI-Powered Features to WhatsApp Business App
