China and AI: a growing security threat?

China is already deploying AI to assist its covert hacking and influence operations overseas, with more direct military threats like cheap autonomous unmanned air or sea vehicles likely to follow later this decade. Neil Thompson reports.

The use of AI in military cyberattacks is likely to be a key focus for countries such the US, as China seeks to leverage its expertise in this area. Credit: Shutterstock/Halfpoint

The US and China are currently in a period of intense economic, scientific, and technological competition, with each hoping to gain dominance in various fields which would provide the more advanced party with powerful productive and military advantages.

Artificial intelligence (AI) is one such technological field, with several applications for actors involved in the global defence sector. These include allowing threat actors to enhance the effectiveness of their cyberwarfare operations against the IT systems of targeted organisations or critical infrastructure; using AI to increase the reach and effectiveness of propaganda and influence operations; and the race to create autonomous AI-controlled weapon systems including aircraft or submarines.

China has begun using AI as a tool in some of these areas and is also interested in stealing the intellectual property of Western entities involved in AI design to improve its homegrown AI capabilities, especially in areas like weapon design. 

Chinese hackers using AI, seeking more

The FBI specifically warns that US entities involved in AI research face significant cybersecurity threats from Chinese-affiliated hacking groups. In October, the heads of the intelligence services participating in the Five Eyes intelligence sharing network — United States, Britain, Canada, Australia and New Zealand — in a Reuters article warned that Chinese hacking efforts to steal foreign intellectual property were being boosted by the attackers’ use of AI tools.

Beijing is therefore deploying AI tools against the West already, as well as designating the technology a priority target for its industrial espionage efforts.

Dakota Cary, a non-resident fellow at the Atlantic Council said there is a division of labour among Chinese state-affiliated hacking groups: “China’s targeting priorities depend on the wants and needs of the CCP and its military. Generally, we see both the military and intelligence services collecting political and military intelligence. The intelligence services and their contractors tend to be the only groups collecting economic intelligence and intellectual property.”

As China’s knowledge of AI deepens and spreads, the number of AI-enhanced cyberattacks targeting non-military targets like Western critical infrastructure providers, defence and technology companies, or academic or public institutions is likely to accelerate in 2024. China’s cyberattacks aim to secure three national security priorities that the ruling Chinese Communist Party (CCP) identifies as China’s “core interests”: keeping the CCP in power, maintaining control over what Beijing sees as Chinese territory, and growing a technology-driven economy.  

AI isn’t going to be redirecting ships any time soon. It’s being implemented in surveillance and misinformation campaigns that are already ongoing, bolstering existing tools instead of revolutionising new ones.

Courtney Manning, a National Security Research Fellow, American Security Project

As a result, in the US, China is mainly targeting large technology companies, defence contractors, and federal IT systems. Targeting a large technology company is certainly easier than attempting to infiltrate the US Government, and, unlike China, the US entrusts third-party contractors with maintaining sensitive data, critical infrastructure, and advanced weapons systems data.

While the power differential of the US and China is quite large, the power differential between China and a third party defence contractor is much narrower, particularly if Chinese hackers can draw upon national AI-research efforts to improve their attacks’ effectiveness.

Chinese information warfare

Chinese influence operations aimed at destabilising adversaries (as opposed to spreading pro-China propaganda) are another emerging area of concern given the recent rapid uptake of generative AI models which can produce authentic-looking content in mediums including audio, video or text, or target and spread messages rapidly via social media and potentially foreign media outlets.

Companies like Google are already producing tools to create AI-generated news articles which digital experts fear could become a vector for spreading misinformation and disinformation. Meanwhile, Microsoft warned earlier this year that Chinese-aligned online actors were using AI-generated visual media since at least March. The AI models used to design online campaigns targeting political divisions in the US over subjects like gun violence and received increased engagement from social media users.

China is also expected to increase its use of AI-generated misinformation and disinformation targeted at the US ahead of the 2024 presidential elections. While not a direct military threat in the same way as a hacker shutting down critical infrastructure remotely, or an AI-guided missile system, sparking government gridlock or political unrest in the US or other Western countries serves China’s strategic goal of keeping its geopolitical adversaries off balance while it develops its economic and military capabilities over the longer term.  

AI military innovation and China

The Pentagon has long claimed a lack of transparency in China's military modernisation defence spending, while the Chinese military’s hacking and surveillance programmes obviously remain undisclosed. This makes it difficult to determine which units in the People’s Liberation Army (PLA) are purchasing and using AI-enabled software and hardware, and what for.

The People’s Liberation Army (and the Chinese Communist Party more broadly) refute any and all accusations of cyber warfare, citing “self-defence” in very narrow circumstances. Naturally, this means no official unit is tasked with carrying out offensive cyber operations on paper.

Courtney Manning, a National Security Research Fellow, American Security Project

The Chinese ground force, air force, and navy are the most transparent Chinese entities with their use of these tools, and most of their purchases are intelligent and autonomous vehicles and predictive algorithms (the same as in the US). The PLA's general capabilities for using AI and machine learning enabled and autonomous weapon systems remains opaque however, even where individual units are known to be integrating these technological developments into their operations.

Some analysts are sceptical that China will soon be able to duplicate in weapon design the early successes it has shown in using AI to generate propaganda or support cyberwarfare attacks below the threshold of military conflicts. China did not demonstrate the use of AI-enhanced systems during its live-fire military exercises around the self-ruled island of Taiwan in 2022 for example, where tensions with the US are rising, continuing to project force using conventional means.

“China [especially prefers] covert activities in the case of Taiwan because the US has such a laser focus on the area and attribution is difficult in the cyber landscape. The increased [Chinese] naval posturing is more a public show of force than a real move to reclaim territory, but AI isn’t going to be redirecting ships any time soon,” said Courtney Manning, a National Security Research Fellow at the non-profit American Security Project.

“It’s being implemented in surveillance and misinformation campaigns that are already ongoing, bolstering existing tools instead of revolutionising new ones,” Manning added.

Nevertheless, automation powered by AI is likely to eventually revolutionise the production and capabilities of ground, air and naval forces later this decade in the same way it has already enhanced cyberwarfare, influence operations and surveillance. China’s defence sector and military establishment have committed significant sums to research, develop, and experiment with machine learning and AI-enhanced weapons systems.

Beijing has already claimed the creation of AI-powered drone swarms. Smart weapon systems are likely to be fielded by the PLA before the end of the decade, as it prepares to meet President Xi Jinping’s aspirations to subdue Taiwan by 2027, widely expected to be the tipping point during which China makes it move to annex the island. 

Australia could be one of the main beneficiaries of this dramatic increase in demand, where private companies and local governments alike are eager to expand the country’s nascent rare earths production. In 2021, Australia produced the fourth-most rare earths in the world. It’s total annual production of 19,958 tonnes remains significantly less than the mammoth 152,407 tonnes produced by China, but a dramatic improvement over the 1,995 tonnes produced domestically in 2011.

The dominance of China in the rare earths space has also encouraged other countries, notably the US, to look further afield for rare earth deposits to diversify their supply of the increasingly vital minerals. With the US eager to ringfence rare earth production within its allies as part of the Inflation Reduction Act, including potentially allowing the Department of Defense to invest in Australian rare earths, there could be an unexpected windfall for Australian rare earths producers.

Credit: US Department of Defense; Department of Energy (originally compiled by Neta Crawford)

Total annual production

$345m: Lynas Rare Earth's planned investment into Mount Weld.

Caption. Credit: 

Phillip Day. Credit: Scotgold Resources