A driver in China got a fine for the smallest possible gesture: scratching his face.
According to the Jilu Evening Post, the driver was only scratching his face — but his gesture looked like he was talking on the phone. An automated camera took a picture of him, and according to Chinese authorities “the traffic surveillance system automatically identifies a driver’s motion and then takes a photo”. Essentially, the AI operating the camera interpreted the gesture as the driver speaking on the phone, and fined him.
The driver, who has only been identified by his surname “Liu” shared the photo on social media, humorously quipping:
“I often see people online exposed for driving and touching [others’] legs,” he said on the popular Sina Weibo microblog,” “but this morning, for touching my face, I was also snapped ‘breaking the rules’!”
After a struggle, he was able to cancel the fine, but this raises important concerns about privacy and AI errors, especially in an “all-seeing” state such as China. The country already has more than 170 million surveillance cameras, with plans to install a further 400 million by 2020. Many of these cameras come with facial recognition technology, and some even have AI capabilities, being able to assess a person’s age, ethnicity, and even gestures. Sometimes, though, they fail.
As the BBC points out, China’s social media was also buzzing with revolt regarding the state’s surveillance policies. China recently implemented a social credit system, intended to standardize the assessment of citizens’ behavior — and input from such cameras are key for the system.
“This is quite embarrassing,” one post commented, “that monitored people have no privacy.”
“Chinese people’s privacy — is that not an important issue?” another asks.
For now, this is indicative of a problem the whole world will have to deal with sooner or later: levels of both AI and surveillance are surging through our society, and we’re still not sure how to deal with them in a way that’s helpful but not intrusive.