China now leads the world in nearly 90% of the “critical technologies” that can significantly boost or endanger national interests, according to the Australian Strategic Policy Institute’s (ASPI) latest Critical Technology Tracker report, underscoring Beijing’s rapid ascent across a wide range of advanced scientific and industrial fields.
In ASPI’s five-year assessment covering 2020 to 2024, China ranked first in 66 out of 74 technologies tracked, including pillars of national power such as nuclear energy, synthetic biology and small satellites. The United States led in the remaining eight areas, notably quantum computing and geoengineering, the report shows. The findings reinforce a pattern identified by ASPI over the past several years: China’s technological capabilities have expanded not only in scale but also in breadth, increasingly spanning both civilian and military-relevant domains.
Published on December 1, the report highlights what it calls “concentrated risk” in several newly added fields where China holds a clear and widening lead. These include cloud and edge computing, computer vision, generative artificial intelligence and grid integration technologies. Some of these areas are rated as carrying a high “technology monopoly risk,” reflecting concerns that dominance by a single country could translate into strategic leverage over supply chains, standards-setting and downstream applications.
Computer vision, a subfield of artificial intelligence that enables machines to interpret images and video, is identified as particularly sensitive because of its dual-use potential in areas ranging from autonomous vehicles and medical diagnostics to surveillance and military targeting. Generative AI and cloud infrastructure, meanwhile, underpin digital ecosystems that shape economic competitiveness and information flows, amplifying their strategic significance.
On the same day the Critical Technology Tracker was released, ASPI published a separate report titled How China’s new AI systems are reshaping human rights, which takes a more normative and political view of Beijing’s technological rise. That study accuses the Chinese Communist Party (CCP) of using large language models (LLMs) and other advanced AI systems to automate censorship, expand surveillance capabilities and pre-emptively suppress dissent.
“China’s extensive AI-powered visual surveillance systems are already well documented,” the report states. “Those practices directly implicate the right to freedom of expression and the right to seek, receive and impart information, including access to accurate contextual information rather than only the absence of prohibited content.”
According to ASPI, Chinese multimodal AI systems—capable of analyzing both text and images—demonstrate consistent patterns of political filtering. Models such as Baidu’s Ernie Bot, Alibaba’s Qwen, Zhipu AI’s GLM and DeepSeek’s VL2 were tested on photographs of politically sensitive events, including the 1989 Tiananmen Square massacre, the 2019 Hong Kong protests and rallies supporting Uyghurs and Tibetans. The report says the models frequently refused to respond, omitted sensitive details or repeated official government narratives.
ASPI attributes these behaviors to Chinese state regulations that require AI systems to adhere to “core socialist values” and to avoid outputs that “harm the national image.” The study argues that such rules effectively embed political controls into the architecture of AI systems, shaping how information is processed and presented.
The report identifies four areas in which the CCP expanded advanced AI use most rapidly between 2023 and 2025: AI-driven censorship of politically sensitive images; integration of AI into policing and the criminal justice system; industrial-scale control and management of online information; and overseas deployment of AI-enabled platforms by Chinese companies. Taken together, ASPI says, these trends indicate that AI is being embedded across governance, security and economic domains to strengthen Beijing’s ability to shape information, behavior and outcomes both domestically and internationally.
The release of ASPI’s findings was quickly followed by a policy response in Canberra. On December 2, the Australian government announced its National AI Plan, which places strong emphasis on security, risk management and resilience as AI capabilities advance.
“The Department of Home Affairs, the National Intelligence Community and law enforcement agencies will continue efforts to mitigate the most serious risks posed by AI proactively,” the plan states. It also stresses that while Australia welcomes global investment to build domestic capability in AI, data infrastructure and energy, some investments may be scrutinized under foreign investment rules to ensure they do not undermine national interest or national security.
ASPI’s warnings about China’s technological rise are not new. In 2023, the institute reported that China led 37 of 44 key technologies, including hypersonic missiles, artificial intelligence, drones and electric batteries, and ranked first in defense- and space-related research. It also identified strong Chinese performance in advanced materials, 5G and 6G telecommunications, clean energy technologies and synthetic biology.
Earlier editions of the tracker showed an even starker long-term shift. During 2003–2007, China led in just three of the technologies ASPI examined. By the 2019–2023 period, it led in 57 of 64. ASPI has repeatedly warned of a high risk that Beijing could secure monopolies in defense-related technologies such as drones, satellites and collaborative robots that can operate alongside humans in industrial and military settings.
Chinese media and commentators, however, have consistently rejected ASPI’s conclusions, accusing the institute of deliberately promoting a “China threat” narrative. Zhang Jingjuan, a columnist at Guancha.cn, described ASPI as “a notorious anti-China think tank” that uses exaggerated claims about China’s technological lead to push the United States, the United Kingdom and Australia toward deeper security cooperation with Japan and South Korea.
She argues that ASPI focuses disproportionately on technologies with military applications—such as radar, satellite navigation and drones—and uses these assessments to justify tighter security alignments under frameworks like AUKUS.
Another commentator, writing under the pen name “Copper Pea” in Zhejiang province, went further, calling Australia’s approach a “manufactured scare campaign.” He accused ASPI of misleading policymakers into boosting defense spending and hardening China policy.
“Public information shows that about 57% of ASPI’s funding comes from defense contractors such as the United States’ Lockheed Martin, Raytheon and the UK’s BAE Systems,” he wrote, arguing that the institute functions more as a mouthpiece for arms manufacturers and government agendas than as a neutral research body. He said Australia faces a structural dilemma, balancing economic dependence on China with security reliance on the United States, and that portraying China as a threat is unfair and counterproductive.
ASPI’s latest reports landed at a sensitive strategic moment. Days later, on December 4, Washington released its National Strategy Report, calling on the Quad—comprising the United States, Japan, India and Australia—to deepen coordination to counter what it described as mounting challenges from China. On December 8, ASPI published another report advocating closer Japan–Australia defense cooperation in the Pacific, including a proposed “partial division of labour” to protect sea lines of communication during a conflict.
That proposal drew fresh criticism from Chinese commentators, who accused ASPI of encouraging Canberra and Tokyo to advance US strategic interests by turning Pacific island nations into arenas of great-power competition. Such critiques reflect the broader geopolitical tension surrounding technology, security and influence in the Indo-Pacific, where assessments of technological leadership are increasingly inseparable from questions of power, values and alignment.
