Will a portion of fried chicken jeopardise your privacy?

There’s a KFC restaurant in the Financial Street area in Beijing that is unique: it is the first smart restaurant in the continent.  How does it work? By using facial recognition, it tries to predict and recall diners’ previous orders. The artificial intelligence system developed with the strategic collaboration of Baidu – otherwise known as the Chinese Google – can recommend suitable food and set meals based on a series of parameters. How? A machine at the store recognises the diner’s face and, based on the diner’s sex, “presumed” age and mood, recommends suitable food which, if not in accordance with the diner’s wishes, can be easily changed by clicking on the alternatives available. Once the order is placed, the diner can pay with his phone and wait for the food to be ready.

For the time being, this is more a publicity stunt than artificial intelligence. «The digitalization of the restaurant will also help to provide faster and easier services» said Zhao Lu, General Manager of the store. How? If the diner visits the store again, the machine that initially took his picture will remember the customer’s dining habits and help to place an order faster.

Consumers have not yet responded enthusiastically to the project. Yet, KFC intends to create an even more customised ordering experience. Quite obviously, apart from the debate over the actual success of this operation, the line between consumer convenience and privacy is very thin when dealing with facial recognition. What will the US chain do with all these data, in a country like China where the idea of privacy is so weak?  The answer is diplomatic, as it can only be in these situations: “Data will not be used for other purposes”. We are not at all sure about this. What is sure is that the experiment, which will soon be replicated in other 5,000 KFC restaurants in China, opens up a rather disturbing future scenario, where the mere ordering of fried chicken will mean jeopardising the privacy of our most personal data.