North Korea may be infamous for being “decades in the past,” but its technology is not nearly so far behind. North Korea has trained a world-class cohort of hackers and IT professionals who rake in billions of dollars for the regime through cryptocurrency theft and remote IT work, even as the vast majority of North Koreans have never accessed the worldwide web. This lopsided technical development means that much of North Korea’s methods for controlling its population still rely on human intelligence and analog systems of identification. While these human-based surveillance systems are thorough, they are also frequently prone to human “error” in the form of bribery and corruption. One shudders to think then, what the North Korean state might be capable of if it were to adopt a Chinese model of AI-enabled omnipotent state surveillance.
A recent report published by US-based think tank 38 North, “Digital Surveillance in North Korea: Moving Toward a Digital Panopticon State,” suggests that North Korea is already experimenting with digitizing state surveillance. The report provides thorough evidence that North Korea has been steadily developing technologies with clear surveillance purposes, including facial recognition and AI-powered CCTV monitoring technologies. This progress raises important questions regarding how soon this sort of technology will be deployable in North Korea and what a tech-powered North Korean surveillance state might look like.
In imagining how North Korea’s surveillance tech might play out on the ground, a new report from Daily NK offers some valuable insight. Journalist Kim Jeong Yoon reported that in 2021, during the height of the COVID-19 pandemic, the North Korean government turned to AI-backed solutions to reinforce controls over domestic travel, ostensibly to prevent the spread of the virus. In April 2021, authorities in South Hamgyong Province installed facial recognition-enabled cameras at gates to multiple train stations in the province to replace manual inspection of travel documents. Similar set-ups were reportedly put into place across the country, including in Pyongyang. In the years since, all of the facial recognition cameras in South Hamgyong Province have fallen into disrepair, but at least three such devices are reportedly still in operation at train gates in the capital city of Pyongyang. (Read the full article here)
Whether these facial recognition devices were domestically produced or imported and modified from China remains unclear. This author believes that the facial recognition software in use is likely of North Korean origin, potentially the Dambo system developed by Kim Il Sung University. Chapter 7 of the 38 North report includes images from a 2019 technological exhibition with a facial recognition-powered access control system for entrance to buildings that could likely be easily adapted for entrance to a train station. The report also highlights domestic coverage of the university’s advanced facial recognition technology dated to December 2021, the same year that Daily NK reports the government installed facial recognition systems at major train stations across the country.
Regardless of software, however, much of the hardware itself is likely of Chinese origin considering the North’s limited capacity for manufacturing highly-integrated electronics. The 38 North report found, for example, that the DT-20 facial recognition access management device advertised by Amnokgang Technology Development Company appears to be produced by the Chinese company “Wit Easy Electronic Co.” Other ostensibly “domestically produced” high-level technologies like North Korean smartphones have been similarly traced back to Chinese manufacturers. North Korean software on Chinese-made hardware or a wholly imported Chinese system would explain why the broken devices in South Hamgyong Province have not yet been replaced or repaired, since repairs would require importing parts or new units from Chinese partners.
This example aligns with three key limiting factors highlighted in the 38 North report: a reliance on imported hardware, insufficient interconnectivity (particularly in more remote areas), and electricity shortages. North Korea may well be working to tackle the second issue (interconnectivity) with its new 4G network, but the country is unlikely to quickly resolve its decades-long struggle to provide a steady flow of electricity. Integrated electronics manufacturing, by a similar token, is likely not an industry that North Korea can rapidly expand on its own.
These hard limits suggest that North Korea is not likely to take the all-seeing CCTV approach favored by richer countries like China and South Korea, but will focus on using “impartial” surveillance to put pressure on the human networks enforcing the regime’s control. When forced by material constraints to target only a small fraction of the population, the most effective strategy is likely to surveil the enforcers themselves. Surveillance of regime-enforcing officials could reduce opportunities for bribery and corruption, which would in turn make crackdowns on “anti-socialist” behaviors more effective. Likewise, facial recognition-backed control over transportation reduces human error, the effectiveness of forged documents, and makes it more difficult for individual officials to take bribes.
Anecdotes from the 38 North report suggest that the North already uses surveillance cameras to keep an eye on enforcement, but that these officials find ways to continue receiving bribes by taking advantage of surveillance blind spots. Considering the fact that Ministry of State Security agents and similar enforcers widely rely on bribes to make a living, and in turn pay part of those bribes to their superiors, this creates incentive for collaboration between enforcement officials to evade surveillance in order to continue generating revenue. Implementation of surveillance cameras may spread around or inflate bribes – such as a percentage of bribes going to officials monitoring CCTV footage – but is unlikely to curb corrupt behavior without complementary policies to reduce financial demands on MSS agents or increase their salaries to the point where corruption poses a greater risk than benefit.
In sum, technology-powered surveillance systems have the potential to further strengthen the regime’s control by reducing opportunities for individuals to bribe their way past enforcers. However, systemic issues like reliance on foreign parts and unreliable electricity will likely relegate tech-based approaches to a minor role for the time being. This author also suspects that the government’s reliance on bribery and corruption to keep enforcers paid will further undermine attempts to effectively implement technological surveillance systems regardless of other technical factors. Likewise, reduced corruption has a ripple effect for ordinary citizens in the form of reduced smuggling opportunities and inflated bribes, further driving up the costs of food and other necessary items and threatening the livelihoods of low-level vendors. Greater surveillance – no matter how imperfect – poses a dire threat to the North Korean people’s lives and livelihoods.
Check out the original 38 North Report here.
Views expressed in this guest column do not necessarily reflect those of Daily NK.
May 20, 2024 at 12:30PM
by DailyNK(North Korean Media)