An eerie image from a restaurant in Oslo, Norway circulated around the web earlier this week. A screen with an advertisement outside a restaurant had apparently crashed, revealing white text on a black background: “Male-Young adult, attention time: 406 out of 406, Smile: 0.” “Female-Young adult, Attention time: 219 out of 219; Glasses: No.” “A crashed advertisement reveals the code of the facial recognition system used by a pizza shop in Oslo,” user Lee Gamble wrote on Twitter. Then, after his tweet got more than 10,000 retweets: “Dystopian HAL pizza broke my Twitter.”
The pizza shop was using an increasingly common advertising tactic: a digital billboard with a hidden camera that scans your face to decide what ad to serve you. In this case, if the facial recognition tech determined you were a man, the billboard would display an ad for a sausage pizza, if it determined you were a woman, it would serve ads for a salad.
A Reddit user identified in Norwegian media as Jeff Newman claimed to have originally taken the photo at Peppes Pizza in Oslo after the ad crashed and revealed a live log of the data it was creating as it scanned his face.
“As I approached the screen to take a picture, the screen began scrolling with my generic information,” Newman wrote on Reddit. “That I am young male, wearing glasses, where I was looking, and if I was smiling and how much I was smiling.” The ad also seems to be tracking how long someone how long someone paid attention to it.
Norwegian Reddit users debated the ethics of the hidden camera. “Yes, it is a breach of privacy,” one wrote. “It doesn't matter that it's a public place. Facial recognition should never be used without the subject's consent.” After Norwegian website Dinside wrote about the secret camera, it was removed.
The use of facial recognition tech in advertisements is more pervasive than most people might think. There are entire companies, like Sightcorp, which serves airports and retailers, that focus on consumer facing facial recognition to optimize advertisements, and Amazon and Microsoft both sell software that analyzes faces to detect a range of emotion. “Understand how your customers feel,” the Sightcorp marketing martial reads. “Detect and measure facial expressions like happiness, surprise, sadness, disgust, anger and fear.” The next step is just figuring out how to use your customers' sadness and fear to sell salads.