KENTUCKY FRIED SURVEILLANCE

KFC and Baidu are applying facial recognition and pseudoscience to predict food orders

At one Chinese KFC, the cashier will tell you what you’re eating. Well, the cashier robot.

The fried chicken chain is partnering with Chinese search giant Baidu to build a new Beijing KFC location that uses artificial intelligence to analyze customers’ faces, and uses traits like age, gender, and facial expression to determine what they want to eat, according to TechCrunch. The restaurant will store the picture of each face, along with the person’s order, so it knows what they ordered next time they’re in the store.

It’s likely that KFC China has data on which age and gender demographics prefer certain foods, and uses the AI’s assessment of the customer’s age and gender to serve those recommendations, rather than performing some calculation on each individual user. A female in her 50s is more likely to be recommended porridge, TechCrunch reports. However, a customer’s mood is also factored into the machine’s decision. (Don’t worry: it’s only a suggestion, and you can override what the AI thinks is best for you.)

As a reminder, predicting behavior based on physical traits is called physiognomy; it has been widely discredited.

While the use of such pseudoscience has the feel of a marketing stunt, Baidu’s facial-recognition technology is industry-leading. By building an image database of customers and their orders, KFC could get a much more deeper understanding of the market.

More importantly, putting this technology in a fast food restaurant normalizes what represents an enormous jump in the data collected by businesses on their customers. An image and order history lets KFC begin to build a detailed dossier on each customer, stored on a machine of unknown security. If other stores implement similar systems, a whole new trove of personal data could be created.

So before you order that eight-piece meal, think about the digital price you’re also paying.

home our picks popular latest obsessions search