Real photos
The Verge recently reported that some photographers have been complaining that some of their photos are being labeled as “made with AI” on Facebook and Instagram.
Several photographers have shared examples over the past few months, with Meta recently marking a photo former White House photographer Pete Souza took of a basketball game as AI-generated. In another recent example, Meta incorrectly added the label to an Instagram photo of the Kolkata Knight Riders winning the Indian Premier League Cricket tournament.
One theory is that the label is applied to photos that are retouched with various Photoshop tools that fill in detail when objects are removed and so on. The article quotes a photographer who raises the appropriate question of where the line between “real” and “altered” should be drawn:
“If ‘retouched’ photos are ‘Made with AI’ then that term effectively has no meaning,” photographer Noah Kalina wrote on Threads. “They might as well auto tag every photograph ‘Not a True Representation of Reality’ if they are serious about protecting people.”
Of course, they are not serious about protecting people. If they were, Facebook would have closed up shop when they started abetting genocides. They are not even that invested in lip-service-level content moderation anymore, as this 404 Media piece details. It’s not clear why they would bother with this sort of “made with AI” labeling, but it should probably be considered more an advertisement for the capabilities of generative models than an effort to protect users.
It is also not clear how this sort of label is protective. It not only represents a sort of moral hazard with respect to media literacy (“don’t worry, the platform will tell you what to think about these images”); it suggests that they only way an image can be harmful is it is not documentary, not “real.” But plenty of nondocumentary images are perfectly harmless, and vice versa; an image’s ability to cause harm doesn’t correspond to the fidelity of its representation. Images, like words, are not true in and of themselves. And images, like words, are potentially harmful not based on how they are created but how they are used, how they are circulated, how they are contextualized. (Imagine Facebook labeling text posts: “These sentences have been perverted by language.”) The harm in what an image depicts is not fully conditioned on it being documentary or unaltered.
Keep reading with a 7-day free trial
Subscribe to Internal exile to keep reading this post and get 7 days of free access to the full post archives.