China spending billions on AI-'intelligent' military technology
The Chinese military, or People’s Liberation Army, is focusing heavily on artificial intelligence. However, China’s race to develop “smarter,” cheaper AI technology for the military is not linear, but instead a many-pronged strategy that involved the central government, domestic companies, and international trade. Gregory Allen of the Center for a New American Security published a report on China’s AI strategy, in which he said:
Chinese military leaders increasingly refer to intelligent or “intelligentized” military technology as their confident expectation for the future basis of warfare. Use of the term “intelligentized” is meant to signify a new stage of military technology beyond the current stage based on information technology.
He also reported that “total Chinese national and local government spending on AI to implement these plans is not publicly disclosed, but it is clearly in the tens of billions of dollars.”
China seems to intend to sell the AI technology it builds, and plans for achieving commercial success in AI-related industries are well underway. According to Jeffrey Ding, the China Lead for the Governance of AI Program at Oxford University, in his Deciphering China’s AI Dream paper, by 2020 China’s core AI industry gross output and AI-related industry gross output may exceed $22.5 billion and $150.8 billion, respectively, making China one of the most advanced countries.
Commercializing its AI output may reap benefits for its military. Allen claimed:
China’s commercial market success has direct relevance to China’s national security, both because it reduces the ability of the United States government to put diplomatic and economic pressure on China and because it increases the technological capabilities available to China’s military and intelligence community.
China’s national AI strategy has strong support from its private sector. We spoke to Irakli Beridze, Head of the Center for Artificial Intelligence and Robotics at UNICRI, and he said:
The type of alliance between government, between the private sector, and academia does bring a unified vision, which they believe can actually yield into positive results for the Chinese economy and also for the development and for the innovation.
China already sees itself as rapidly closing the gap between itself and the US when it comes to AI. That said, according to Allen:
China’s success has been enabled by its access to global technology research and markets. Many seemingly “Chinese” AI achievements are actually achievements of multinational research teams and companies, and such international collaboration has been critical to China’s research progress.
China’s aggressive pursuit of an “intelligent military” has analysts fearing that it could trigger an AI arms race, involving not only the US, but also its neighboring countries, particularly Japan. Allen stated that the “increased use of AI systems would make misperceptions and unintentional conflict escalation more likely due to the lack of well-defined norms regarding the use of such systems.”
That said, China still has a ways to go in overcoming inherent problems with China’s AI ecosystem. Among these are a lack of AI talent, lag in core technologies, funding inequalities, market competition, arms control concerns, and current external dependencies.
What exactly does the Chinese military have in the works with AI technology? In this article, we discuss three AI use-cases in China’s military:
- Autonomous vehicles
- Facial recognition
- Situational Awareness
Autonomous Vehicles and Drones
According to The National Interest, China is the biggest exporter of unmanned combat aerial vehicles (UCAVs), otherwise known as combat drones. The Stockholm International Peace Research Institute (SIPRI) reports:
There is widespread discussion about the impact of UCAV proliferation on peace and security. China has become the primary exporter of UCAVs. Whereas China exported 10 UCAVs to 2 countries in 2009-13, in 2014-18 it exported 153 to 13 countries—5 of which are in the Middle East: Egypt, Iraq, Jordan, Saudi Arabia and the United Arab Emirates (UAE).
Ziyan UAV, which produces the attack helicopter drone Blowfish A2, demonstrated its capabilities to interested buyers in Malaysia at the 15th Langkawi International Maritime and Aerospace Exhibition (LIMA). It has already sold the drone to the UAE and is in negotiations with Pakistan and Saudi Arabia. Below is a one-minute video showing how the Blowfish A2 uses computer vision to identify targets:
The US has always been at the forefront of drone technology but has restrictive policies that prevent it from exporting drones wholesale to other countries. Even if the US sells them, it restricts the capabilities of the drones for approved uses.
China has no such policies or restrictions, and it is offering its combat drones at the lowest prices.
The Wing Loong I-D is the latest in the Wing Loong Unmanned Aircraft System series developed by the Chengdu Aircraft Design & Research Institute.
Below is a 46-second video showing it during its maiden flight:
China also has autonomous vehicles its military uses itself. One of these is the Marine Lizard, which is not a UCAV, but an autonomous amphibious landing vehicle. According to the state-owned developer China Shipbuilding Industry Corporation, it can “plot out its own route, swim to shore, avoid obstacles, and it can also be remotely controlled by an operator.” Given recent forays into expanding its territory in the South Pacific, the Marine Lizard ay prove useful to their aims if it works as advertised.
Growing concerns over the unbridled use of AI in the Chinese military appears to have substantial merit. However, it has not stopped private US companies from collaborating with Chinese researchers to develop AI-based technology.
One of these is Microsoft, which is working with military-funded National University of Defence Technology (NUDT) on surveillance technology. Another is Google, which opened an AI research facility in Beijing in 2017. The lab develops AI and machine learning tools, including the open source platform TensorFlow. These companies disclaim any responsibility for the eventual use of their technology down the line. Former Google Cloud CEO Diane Greene stated, “we can not control all downstream uses of our technology.”
That is precisely the issue with this collaboration with China. While the results of this type of research are available to the public, China can use these technologies with far more freedom than countries like the US can. For example, facial recognition technology from top AI companies SenseTime and Megvii is in wide use by local law enforcement in some parts of China, particularly Xingjian, Shaanxi, and Shenzhen.
Below is a one-minute video showing how facial recognition technology in Shenzhen works in deterring jaywalking:
While authorities claim AI surveillance resulted in higher success in preventing jaywalking, finding missing persons and apprehending terrorists, there is a big potential for misuse, such as in targeting minority groups. Eventually using the technology to target military persons of interest is highly probable.
There is also the question of privacy concerns with the use of this technology. Facial recognition company SenseNets inadvertently released personal information on 2.5 million Chinese citizens it was tracking via an unsecured database.
The biggest issue with facial recognition technology in a national defense sense, however, is the export of these capabilities to governments that want to control their own populations using the same methods. Over time, this may make this type of control through surveillance not only acceptable but the norm in many countries.
China has the capacity to gather huge amounts of data from civilian as well as military operations because of its much-vaunted academe-government-private sector cooperation. However, data access means nothing without analysis, especially for situational awareness.
In a move to improve its situational awareness, China purchased war-room mapping software from Belgian company Luciad. In particular, it purchased the data analytics program LuciadLightspeed, which according to the site “allows for on-the-fly combination of any data with accurate geospatial transformations and can perform advanced functions including real-time video draping, immediate line-of-sight analysis, and dynamic density calculations.”
Below is a 2-minute video demonstrating the various uses of LuciadLightspeed:
The software package would allow the Chinese military to use the data it gathered from its drones, facial recognition technology, sensors, satellites, and radar to visualize and analyze enemy movements and other information. This is the same software used by the US military and the North Atlantic Treaty Organisation (NATO).
The Race to 5G
A related issue to data access and analysis is data transmission. While developing and acquiring powerful AI software and algorithms are definitely priorities in China’s military AI strategy, establishing a fast and reliable way to transmit the data and connect devices may be even more important.
According to former US Air Force Brigadier General Robert Spalding, China could “weaponize cities” if it succeeds in dominating the fifth generation (5G) of mobile communications. According to a public statement released by former military leaders of the US military, “Chinese-designed 5G networks will provide near-persistent data transfer back to China that the Chinese government could capture at will.” This is the main reason there is such an intense competition to establish a stable 5G network, despite expert opinion that this infrastructure is more vulnerable to hacking.
It is likely that the race to dominate the 5G network will have some profound consequences for the use of AI in any military.
The use of AI in the Chinese military reveals the uncertainty and disruptiveness of the technology. With AI changing the rules of the game, conventional warfare may have little place on today’s battlefield, and China is taking full advantage. Because there is little oversight over the research and development of AI, China is operating in a gray area that other countries may find hard to mimic.
According to Allen, China is the leader in many aspects related to AI, including research papers, number of patents, and venture capital. However, it is interesting to note that this supremacy is largely due to access to foreign funding and technology research.
Additionally, the AI sword can cut both ways. Much is yet unknown about AI and machine learning technology, which is the reason many developed countries are proceeding with caution when adopting it, especially in the military. The rapid pursuit of China for AI supremacy, with little time for testing for accuracy and reliability, can backfire quite badly.