I’ve seen them following me around on Facebook: photographs of older couples celebrating a 75th wedding anniversary, or children carving astonishing things out of wood, or beautiful stewardesses donning uniforms from decades past.

Granted, this is all pretty standard fare for social media. What’s new, however, is that upon closer inspection, all of these images appear to be AI-generated slop.

What makes me so sure? Thankfully, AI-generated photographs still have a lot of “tells.” This week, we hear from Kellogg’s Matt Groh, along with collaborators Negar Kamali and Karyn Nakamura, on some of the most prevalent giveaways.

Functional implausibilities

You may have heard the advice to always count the fingers. (Which is still a good strategy!) But beyond anatomical anomalies, the researchers suggest looking at the way objects operate and interact. After all, these text-to-image AI models don’t actually know how things work in the real world, leaving plenty of room for things to go … amiss.

A college-logo sweatshirt might have the institution’s name misspelled—or be written in an unconventional font. A diner could be sitting at a table with his hand inside of a hamburger. A tennis racquet’s strings may hang slack. A normally floppy slice of pizza could be sticking straight out.

The text on this person’s sweatshirt offers a clue to it being an AI-generated image. Photo courtesy of Matt Groh.

“More-complex scenes introduce more opportunities for artifacts and implausibilities—and offer additional context that can aid in detection,” Kamali says. “Group photos, for instance, tend to reveal more inconsistencies.”

Zoom in on details of objects, such as buttons, watches, or buckles. Groh says funky backpacks are a classic artifact.

“When there’s interactions between people and objects, there are often things that don’t look quite right,” Groh says. “Sometimes a backpack strap just looks like it merges into your shirt. I’ve never seen a backpack strap that does that. Also, I think your backpack would fall off your back.”

For more tips on recognizing AI-generated images (and several more examples, too), visit Kellogg Insight.

Jessica Love, editor in chief
Kellogg Insight