Meta announced in February that it would start labeling photos created with AI tools on its social networks. Since May, Meta has been regularly adding the label “Made with AI” to some photos on Facebook, Instagram, and Threads. However, the company’s approach has drawn criticism from users and photographers for labeling photos that were not created with AI tools.

Instagram labels real photos as ‘Made with AI’

One example of Instagram adding the AI label was a photo of the Kolkata Knight Riders winning the Indian Premier League cricket tournament. This label only appears on mobile apps, not on the web. Many photographers found this label incorrectly added to their photos. Former White House photographer Pete Souza noted in an Instagram post that his photo was labeled with this new tag. Souza believes that using Adobe’s cropping tool and saving the image as a JPEG may have triggered Instagram’s algorithm to add this label.

Instagram has not made an official statement regarding users’ experiences or other photographers’ posts. In its February blog post, Meta stated that it detects labels using image metadata. Meta also mentioned developing advanced tools to tag images created with tools from companies like Google, OpenAI, Microsoft, Adobe, Midjourney, and Shutterstock by adding metadata.

According to PetaPixel’s report last week, Instagram appears to be adding the “Made with AI” label when photographers use tools like Adobe’s Generative AI Fill to remove objects. While it is unclear when Meta automatically adds this label, some photographers support Meta’s approach, arguing that any use of AI tools should be disclosed.

Currently, Instagram does not provide separate labels to indicate whether a photographer used a tool to clean up a photo or create it with AI. It can be difficult for users to understand how much AI was used in a photo. Meta’s label states, “Generative AI may have been used to create or edit content in this post,” but you need to tap on the label to see this.

Despite this approach, many photos clearly generated by AI on Meta’s platforms remain untagged by the algorithm. With the US elections just a few months away, social media companies are under more pressure than ever to accurately manage AI-generated content.

What do you think? Share your thoughts in the comments below.