Has China Taken It Too Far With Surveillance Tech?

July 19, 2018 - 8 minutes read

AI app developerFacial recognition has come a long way in the past few years. But there’s still progress to be made in terms of eliminating biases and other flaws. Regardless of this fact, China is already employing the technology to track its own citizens.

This police-state type of surveillance has managed to cook up quite a bit of controversy and criticism since it first broke news.

2018 or 1984?

Beijing is China’s, if not the world’s, biggest tech hub. It’s set to overshadow San Francisco in terms of tech; the city is attracting a huge number of artificial intelligence (AI) developers and smart city architects from around the world. And it’s on track to become one of the world’s first smart cities. So it’s not really that big of a surprise that it would implement AI-powered facial recognition to track residents and catch criminals, right?

The police obviously like this technology for easing their jobs as the Chinese population continues to rapidly grow. “In the past, it was all about instinct. If you missed something, you missed it,” says Shan Jun, the deputy chief of the police at a railway station in Zhengzhou, referring to catching a criminal on the streets.

AI app developer

The country has nearly 200 million cameras dedicated to surveilling, whereas the U.S. has only a quarter of that amount. That number will grow to 300 million by 2020. The government aims to connect location data with Internet use, phone communications, hotel stays, and trips on planes, trains, and cars. But although this technology might help clean up the streets and keep children safe, it’s also being used to control citizens and scare them into behaving well.

Who’s Powering the Progress?

Chinese startups are the main fuel behind this growth and technological progress. These companies are aiding the government in building a nationwide surveillance network that an official can access at any time. And many of them are using every avenue they can; quite a few of them encourage their employees to utilize their technology in-office so it can be optimized.

Skynet, the country’s big data network for the police, is flooded with millions of facial images every day from startups who openly share the data they’re collecting with the platform.

AI app developer

Globally, many of the startups compete in international competitions to flex their muscles and show others how far they’ve gotten. Yitu, an AI and computer vision startup, landed the first place prize in 2017’s U.S. Office of the Director of National Intelligence competition for their facial recognition algorithms. Many other Chinese startups also ranked highly in this contest.

Embarrassing Citizens Into Compliance

Like many technologies, facial recognition’s reputation largely depends on how it’s being used. And China’s implementations don’t exactly cast it in a good light.

If you owe a debt, you might see your name and government I.D. number on a public TV in China. Or maybe you’re an upright citizen, but you had to jaywalk across the street so you could get to the office on time. That’ll land you on the public TV too. Embarrassing, right? China hopes this type of public intimidation will decrease bad or illegal behaviors.

Guan Yue, a spokesperson, explains, “If you are captured by the system and you don’t see it, your neighbors or colleagues will, and they will gossip about it. That’s too embarrassing for people to take.” Using data previously collected elsewhere in the city about the person’s facial structure, their walking pattern, and their clothing, it’s nothing short of a public humiliation for most people who get caught red-handed.

AI app developer

Martin Chorzempa is a fellow at Washington DC’s Peterson Institute for International Economics. He says, “The whole point is that people don’t know if they’re being monitored, and that uncertainty makes people more obedient. This is potentially a totally new way for the government to manage the economy and society. The goal is algorithmic governance.”

The All-Seeing Eyes of Justice

Zhengzhou police recently got to play with some high-tech specs when they put on facial recognition glasses to identify passengers in a train station. The glasses aren’t perfect yet; they only work well if the person-to-be-identified stands still for a few seconds.

That’s not realistic in a train station, but it didn’t stop police from apprehending a criminal who was wanted to smuggling heroin. The officers detailed what happened when they brought the suspected heroin dealer to the police station. Jun says his deputies let the criminal know about the glasses and said that no matter what the suspect tried to say, it didn’t matter.

“Because he was afraid of being found out by the advanced technology, he confessed,” he says. The suspect, by that time, had already ingested 60 packs of heroin in an attempt to cover up the crime. But Jun says, “We didn’t even use any interrogation techniques. He simply gave it all up.”

AI app developer

The country has a nationwide database of 20 to 30 million people who are suspected criminals, terrorists, drug dealers and traffickers, political activists, and more. The sheer size of this single database is still too big for our computers to sift through seamlessly. As a result, humans are still parsing through images, making the current system a bit of a clunky technology.

Unforeseen Consequences

The Chinese government is telling its citizens that they should believe they are being watched with incredibly powerful surveillance technology, not some patched up AI. China’s police alone plan to spend another $30 billion on portable technology and surveillance-enabled wearables in the next few years.

We can’t know what the far-reaching implications of China’s fast adoption of the surveillance technology will be. But one thing’s for sure — it won’t be long before other countries, like the United States, have to make a decision about how it utilizes AI to improve surveillance.

It’s a tough scenario to ponder, but one we should all consider.

Tags: , , , , , , , , , , , , , , , , , , , , , ,