US-China Artificial Intelligence Race
Geopolitical Implications
In the intricate tapestry of 21st-century global dynamics, a profound transformation is unfolding, redefining the very nature of international rivalry. What once centred on territorial conflicts, military alliances, and economic sanctions has now shifted decisively into the realm of technology, with artificial intelligence (AI) emerging as the ultimate strategic asset and a defining force.
This shift has inaugurated a new chapter in global power dynamics, marked by an intensifying competition between the United States (US) and China for AI supremacy. This rivalry extends far beyond mere technological advancement, influencing national security, economic policies, social governance, and even the fundamental notion of truth. The term “AI Cold War,” first appearing in 2018, encapsulates this geopolitical tension, suggesting a new era of competition waged in AI technology rather than nuclear capabilities or ideology, becoming a self-fulfilling prophecy that fuels an arms race.
This article sheds light on the high-stakes US-China AI Race, exploring its various battlefields, geopolitical ripples, military implications, and AI’s concerning influence on disinformation, ultimately examining the scenarios and global impact this defining contest holds for the future.
The Stakes of AI Supremacy
The current landscape of AI competition carries an urgency often compared to a “Sputnik moment,” where rapid advancements by one power necessitate a swift, strategic response from others. The launch of OpenAI’s ChatGPT in November 2022, a generative AI chatbot, vividly illustrated AI’s capacity to transform nearly every facet of economic and social life, further intensifying the AI race.
This event, coupled with the Biden administration’s sweeping semiconductor export controls a month prior, marked the inauguration of a new era of US-China innovation competition. China’s 2017 AI Development Plan, which laid out a strategy to become the global leader in AI by 2030, was a pivotal moment that cemented the “AI Cold War” narrative.
AI supremacy in this context means achieving global leadership across multiple critical domains, moving beyond traditional measures of power. It signifies the ability to assert dominance on the world stage, strengthen military capabilities, and accelerate economic growth.
Nations excelling in AI gain a strategic advantage in national security, encompassing autonomous warfare and sophisticated cyber defence. Economically, AI promises to revolutionise national industries, enhance decision-making on a policy level, and stimulate innovation, creating entirely new markets and boosting global competitiveness. Moreover, leadership in AI confers the power to set global standards and norms, influencing policymaking and shaping the future of international relations.
This race is existential because AI is no longer just an innovation but is considered the “ultimate strategic asset” of the 21st century. The winner of this contest is poised to gain enduring economic and geopolitical advantages. The dual-use nature of AI, with both civilian and military applications, raises the stakes significantly, heightening the risk of miscalculation and unintended escalation that could potentially lead to conflict.
The proliferation of AI fundamentally undermines global governance systems, human rights frameworks, and mutual confidence between citizens and states due to the malicious use of AI-generated content like deepfakes. The potential for AI-enabled biological, chemical, cyber, or even nuclear weapons poses catastrophic or existential risks to humanity.
As AI development becomes increasingly linked to national security and economic strategy, its influence extends beyond the superpowers, compelling nations worldwide to adapt to an uncertain international landscape.
Current Battlefield: Key AI Sectors
The competition for AI supremacy is playing out across several interconnected sectors, each representing a critical battleground.
Semiconductor Wars
At the very core of AI innovation lies hardware: specialised processors known as AI chips, which are rooted in physical materials and minerals, not merely software. The production of these advanced AI chips reveals a deep interdependence on global semiconductor supply chains. Essential base materials such as high-purity silicon, aluminium, copper, gallium, germanium, palladium, and rare earth elements are crucial.
China controls a dominant share of the market for high-purity alumina (HPA), critical for chip insulation, while advanced silicon refining capabilities are primarily in Japan, South Korea, and China. The fabrication of leading-edge AI chips is complex and capital-intensive, concentrated in East Asia, most notably Taiwan’s TSMC, which manufactures for major American companies like Nvidia, AMD, and Apple. Taiwan’s critical role in this supply chain is a significant vulnerability, with its non-recognition of sovereignty by China creating a “semiconductor ‘cold war’”.
To maintain its technological dominance and safeguard national security interests, the US has imposed stringent export controls and semiconductor tariffs on advanced AI chip technology destined for China. These restrictions extend to critical manufacturing equipment, such as extreme ultraviolet (EUV) lithography tools, essential for next-generation semiconductors. While these measures aim to throttle China’s advancements, they have also spurred China to intensify its drive for technological self-reliance. Chinese tech giants like Huawei and Baidu are rapidly advancing their AI chip development, although they still rely on certain US semiconductor technologies for optimal performance.
However, Huawei has already unveiled the seven-nanometer Kirin 9000 and Ascend 910 AI processor series, showcasing ingenuity despite US restrictions. China has also significantly increased its spending on older generation lithography tools, fueling a dramatic expansion of its semiconductor sector, which is projected to account for approximately one-third of global production in legacy chips.
DeepSeek’s success in developing a highly competitive open-source Large Language Model (LLM) under US restrictions further suggests that these controls may inadvertently accelerate Chinese innovation by forcing efficiency improvements.
Europe is also charting a unique path, focusing on “chip sovereignty” to reduce its dependency on Asian chipmakers and secure technological independence. Through initiatives like the EU Chips Act, backed by substantial state funding (30 billion euros), Europe aims to build indigenous chip design and fabrication capabilities and attract foreign semiconductor investment.
Germany is set to host one of the most advanced semiconductor fabrication plants outside the US, bolstering Europe’s role in AI chip manufacturing. However, the scale of these initiatives still falls short compared to Asia’s massive manufacturing capacity, and concerns about protectionism and a “subsidies ‘race to the bottom’” have been raised.
Data and Algorithms
Beyond hardware, the battle for AI dominance is fiercely contested in the realms of data and algorithms. China possesses a vast pool of researchers and data, which it leverages alongside a growing ecosystem of innovative companies like Huawei and Baidu. Its centralised state support, with direct state funding for AI projects and national computing centres, has been instrumental in its technological ascent. The US, in contrast, benefits from a dynamic innovation ecosystem with world-class research institutions, a strong culture of innovation and entrepreneurship, and significant private sector investment from tech giants like Google, Meta, and Nvidia.
China has notably outpaced the US in AI and machine learning (ML) patents every year since 2021, granting more than double the US patents in 2023 alone. This trend is concerning because patents are seen as the “seeds of innovation” and a leading indicator of future technological prowess.
However, China also faces vulnerabilities. Its strict censorship protocols could lead to data deficits and biased data sources for training LLMs, as a heavily censored internet may not provide the diverse and unbiased information needed for optimal AI model development. Additionally, China’s market and economic systems are considered less efficient in leveraging government-funded research and development (R&D) investments compared to the US. China also lags behind the US in the diffusion rates of many digital technologies across industrial applications, such as smart sensors and cloud computing.
Talent and Research
The “AI Cold War” is also a fierce battle for human capital and intellectual leadership. Historically, the US has excelled at attracting top STEM (science, technology, engineering, and mathematics) talent from abroad, including a significant number of Chinese researchers, which has been a key strength of its innovation ecosystem. World-class research institutions like those in Silicon Valley, along with a culture of innovation, have fostered this talent pool.
However, the US risks undermining its AI leadership due to what some describe as an “unfavourable talent policy,” including slashed funding for government labs and universities, and visa restrictions on Chinese international students. Donald Trump’s administration’s sudden revocation and reinstatement of 1,800 student visas in April 2025, and efforts to aggressively revoke visas for Chinese students in critical fields, are seen as measures that could deter current and future foreign students from pursuing opportunities in the US, thereby impacting AI innovation.
In contrast, China is rapidly advancing its AI talent cultivation efforts and actively aims to lure foreign scientists. China generates four times as many STEM graduates annually as the US, providing a vast pool of researchers and data. Chinese universities like Tsinghua and Peking are emerging as global hubs for AI research, competing with leading US institutions in producing cutting-edge AI research.
The CCP’s broader strategy includes heavy investment in education and talent programs to attract international experts, viewing this as crucial for achieving its ambition of becoming a global AI leader by 2030. This divergence in talent policy, with China actively cultivating and attracting global talent while the US faces potential talent drain, could have significant impacts on the American talent pipeline and its leadership in STEM and AI research.
Geopolitical Ripples: Shifting Alliances
The escalating competition between the US and China in AI is not merely a bilateral struggle but a transformative force reshaping the global order, with profound geopolitical ripples extending far beyond the two major powers.
Tech Blocs Forming
One of the most striking consequences of this rivalry is the fragmentation or “polarisation of the global landscape”. Smaller nations often find themselves caught in the middle, navigating the complex challenges of balancing ties with both superpowers.
Many are compelled to align with one side over the other, driven by economic dependencies, technological aspirations, or strategic needs.
This alignment is causing the global AI ecosystem to fragment as nations gravitate toward distinct spheres of influence, potentially leading to the existence of two parallel digital ecosystems.
This dynamic encourages countries that consider themselves democracies to align their technological ecosystems with that of the US in a process labeled “re-globalization”.
Neutral Nations Caught in the Middle
While the US and China push for technological dominance, other countries are also asserting their own technological sovereignty. Europe, in particular, is charting a “unique path” in the AI semiconductor race, focusing on reducing dependency on Asian chipmakers rather than engaging in a direct trade confrontation like the US and China.
Through initiatives like the EU Chips Act and substantial state-backed funding, Europe aims to build indigenous chip design and fabrication capabilities and attract foreign semiconductor investment. However, despite these efforts, the scale of EU initiatives still falls short compared to the massive manufacturing capacity of Asia’s chip giants.
Beyond the two superpowers and Europe, various other countries like the UK, Canada, France, Singapore, India, South Korea, and Israel have recognised AI’s potential and are positioning themselves as key players in the pursuit of AI supremacy.
Russia has also aspired to be a leader in AI, attempting to use it to influence operations in the West, but its advanced initiatives have been impeded by the conflict in Ukraine and US sanctions. This underscores how the AI race is altering the global power structure, pushing smaller nations to make strategic decisions that will impact the future of international relations.
The Race for AI Standards
The competition for AI leadership is increasingly turning to the domain of global governance and standards. Both the US and China acknowledge the importance of international cooperation for AI safety and risk management.
China has launched its “Global AI Governance Initiative,” with outreach to the Global South, to ensure its voice remains influential in AI governance. This initiative promotes tiered testing systems based on AI risk levels and supports discussions within the UN framework to establish an international institution for AI governance. China also emphasises equal access to AI technology across global borders and fairness in AI governance.
The US seeks to lead an inclusive international network of AI safety institutes and promote common AI standards with like-minded countries, aiming to foster commonly accepted scientific methodologies and develop shared AI safety evaluations.
The Bletchley Declaration, signed by the US, China, and 27 other nations in November 2023, is a notable joint statement acknowledging safety risks at the frontier of AI, particularly from misuse or unintended issues of control.
However, new political trends of “neo-nationalism” and national security/commercial interests may hinder multilateral efforts for AI governance, as technologically advanced member states may see little incentive in letting international forums govern what they perceive as their own lucrative and proprietary technology.
Military and Security Implications
AI’s transformative power extends profoundly into the military and security domains, fundamentally reshaping modern warfare and international stability.
Autonomous Weapon Systems
Both the US and China are heavily integrating AI into their defence systems, viewing it as a crucial factor in enhancing their military capabilities and securing a competitive advantage.
This includes the development and deployment of AI-driven autonomous systems such as drones for surveillance and tactical operations, robotics for logistics and combat, and smart mines for naval operations.
The ongoing Russian-Ukrainian conflict provides a striking example of AI’s impact, where Ukraine’s use of cutting-edge technology, often in the form of AI-enhanced drones provided by Western nations, has enabled it to hold its ground against its larger rival. This demonstrates how asymmetric battles fueled by advanced technologies can level the playing field in conflicts.
China’s People’s Liberation Army (PLA) has adopted an “intelligentized warfare” doctrine, aiming to integrate AI and autonomy across all domains of conflict to gain a crucial systemic military edge.
China has focused heavily on developing and investing in AI-driven autonomous systems, including advanced drones for surveillance and combat missions, and robotic platforms to enhance battlefield operations and decision-making processes.
Analysts estimate China’s annual military expenditure on AI to be in the “low billions” of US dollars, comparable to the Pentagon’s efforts. China has gone a step further by incorporating AI into its naval operations in the South China Sea, utilising autonomous submarines, smart mines, and surveillance drones to observe and potentially counter foreign military actions.
Cyber Warfare and Intelligence
AI significantly enhances capabilities in cyber warfare and intelligence gathering. AI systems emerge as a frontline defender against numerous waves of cyberattacks, with the capacity to identify, anticipate, and eliminate these threats, thereby enhancing states’ defence strategies.
In the near term (over the next five years), AI is expected to primarily enter the military picture in three critical areas: intelligence, data processing and battle management, and the coordination of robotic swarms. For intelligence, AI’s raw processing power is enormously helpful for handling vast amounts of data, especially for tracking movements and locations across the Earth’s surface multiple times a day, a task beyond human capacity by traditional means.
In battle management during intense periods of future conflict, AI is indispensable for figuring out, in real time, how to defend against large-scale barrages involving hundreds or even thousands of missiles and drones, such as those Iran launched at Israel or that China might launch at Taiwan.
Coordinating defences and allocating interceptors against such attacks would likely be beyond the capacities of human battle managers. The US “Replicator” initiative, designed to field large numbers of uninhabited robotic systems (sea, air, ground), will rely heavily on AI for various missions, building on lessons from the Ukraine war.
AI also has the potential to empower individuals to create bioweapons and conduct cyberattacks, lowering the barriers for terrorist groups to execute large-scale attacks previously limited to nation-states.
Deterrence and Escalation
The accelerating competition in AI, a dual-use technology, without adequate cooperation or comprehensive regulation, heightens the risk of miscalculation and unintended escalation, potentially leading to armed conflict. The deployment of sophisticated and potentially lethal AI-driven weapons poses a serious threat to global stability, as AI in military contexts can lead to unpredictable outcomes and misinterpretations.
However, there is also a recognition of the severe dangers posed by AI. In a rare moment of alignment, the presidents of the US and China released a joint statement pledging “to maintain human control over the decision to use nuclear weapons”. This act, although terse, was striking, indicating a mutual desire to establish a baseline of stability in a space without custom. Experts from both countries express concern about “extreme risks” from advanced AI, such as “intelligence explosions,” AI systems that self-replicate or deceive, and those that escape human control or develop malicious intent. These concerns include the potential for AI-enabled biological, chemical, cyber, or nuclear weapons that could cause catastrophic or even existential risks to humanity.
Beyond the US-China rivalry, both powers face a shared and growing threat from rogue states and non-state actors (e.g., terrorist groups, lone wolves) who may be more willing to exploit AI without restraint. Unlike nuclear technology, which requires enriched uranium and complex infrastructure, AI is “far more accessible and significantly harder to monitor,” making it a “wild card” that could be weaponised to scale destructive capabilities far beyond what is currently possible.
This shared vulnerability should compel both powers to act in parallel, developing new multilateral coalitions and mechanisms like export control frameworks or monitoring regimes.
AI's Dark Side: The Influence on Disinformation
Artificial intelligence significantly influences disinformation by enabling the creation and spread of deceptive content, posing serious threats to global stability, security, and the very notion of truth. The proliferation of AI-generated content corrodes the notion of truth and mutual confidence between citizens and their states, as well as among states. This disruption can fundamentally undermine global governance systems and human rights frameworks.
Sophisticated Deepfakes
AI allows for the development of advanced algorithmic programs that can easily alter or manufacture audio and video images, creating impersonations that are virtually identical to the original.
Deep-learning facial recognition algorithms can accurately copy eye-motion, simulate facial expressions, and even synthesise speech by analysing breathing patterns in combination with tongue and lip movements.These AI-generated deepfakes are used maliciously to discredit individuals, misrepresent their positions on key policy issues, or create “Hollywood-effect propaganda images and videos”.
For instance, the People’s Republic of China (PRC) has created AI-generated deepfakes of US politicians to discredit or misrepresent their positions on key policy issues. Such fake videos could trigger public panic (e.g., non-existent epidemics), mass demonstrations (e.g., high-ranking officials in bribing scenes), or forged security incidents that may provoke serious international escalations.
Amplification of Foreign Interference and Propaganda
AI augments foreign powers’ ability to go on the “offensive” and drive wedges into voter bases, as seen with the PRC and other foreign actors targeting the US. Foreign governments can leverage AI to create disinformation campaigns that depict a nation as violence-ridden or fabricate security incidents that could provoke international escalation.
The Canadian Centre for Cyber Security has warned that the PRC and Russia will continue to be responsible for most attributed AI-enabled cyber threats and disinformation activities targeting democratic processes.
This form of AI-augmented disinformation meets the criteria for a higher threshold threat on the continuum between influence and interference.
Notably, Beijing’s experiments with AI-enabled electoral interference in the US have been restrained compared to Russia’s efforts, but its meddling in Canada’s 2019 and 2021 federal elections, though primarily non-AI generated, hints at the potential for such activities.
Erosion of Truth and Trust
The proliferation of AI-generated content, including deepfakes and polarising contents, corrodes the very notion of truth and mutual confidence between citizens and their states, as well as among states. This disruption can fundamentally undermine global governance systems and human rights frameworks.
By distorting reality and manufacturing falsehoods, AI-powered disinformation erodes the collective consciousness’s ability to discern truth, fostering an information environment that breeds mistrust in national governance and global stewardship. This can lead to popular revolts and long-lasting detrimental implications for international forums dealing with peace and security.
Manipulation of Human Behaviour
AI systems, especially when coupled with technologies like the Internet of Things (IoT) and biotech data, are increasingly programmed to predict and channel human behaviour in manipulative ways. These – still unregulated – AI technologies are increasingly able to channel behavioural and biological data in novel and manipulative ways.
This can include influencing consumer choices, assessing job applicants, and potentially monitoring populations via biometric data, raising significant privacy and human rights concerns. For instance, AI-powered systems are already capable of accurate and speedy analytics of urban traffic patterns, financial markets, and consumer behaviour.
If AI can penetrate and steer individual and group human behaviours, it inevitably disrupts the very notion of human rights as embedded in the UN Human Rights Charter, as well as principles of peaceful coexistence, security, prosperity, and equality among states. This necessitates a re-evaluation of frameworks for monitoring and implementing human rights in a digitally controlled society.
The Future: Scenarios and Global Impact
The trajectory of the AI race presents a critical juncture, with potential paths toward either increased cooperation or heightened confrontation, each with profound implications for global stability and human well-being.
Cooperation or Confrontation
There are varied perspectives on the nature of the US-China AI rivalry. Some, including former US officials, frame it as an “AI arms race” for dominance, where the winner gains enduring advantages, often justifying aggressive policies like export controls. Vice President JD Vance explicitly called competition on AI with China an “arms race”.
Others, like Ryan Hass, argue that the US and China will be “running side-by-side” in AI development, with neither gaining a decisive edge. This view contends that US export controls have not throttled China’s progress but rather galvanised its self-reliance and innovation, as exemplified by DeepSeek’s success under restrictions. The performance gap between the best Chinese and US AI models has significantly shrunk, suggesting a “new normal” of parallel advancement.
The effectiveness of export controls is debated. While they aim to impede China’s ability to advance its AI capabilities by limiting access to advanced chips and critical manufacturing equipment, they can also galvanise China’s efforts toward technological self-reliance and efficient innovation.
DeepSeek’s innovation under constraints suggests that progress can indeed come from constraint. Policymakers are urged to reorient US policy to focus on strengthening America rather than solely trying to weaken China, investing wisely in industry, infrastructure, and talent.
Despite intense competition, both the US and China have expressed interest in global cooperation on AI safety and risk management.
The Bletchley Declaration and International Dialogues on AI Safety highlight a mutual acknowledgment of extreme AI risks and the need for coordinated global action.
A more nuanced policy that restricts military applications of AI while allowing for open collaboration in non-military research, grounded in transparency and international standards, could better balance competition with cooperation. Addressing shared threats from rogue actors could compel both powers to act in parallel, developing new multilateral mechanisms for export control frameworks or monitoring regimes.
Impact on Developing Nations
The AI race also carries significant implications for developing nations, particularly the Global South. There is a risk of “cyber-colonisation,” where leading AI nations or technology platforms extract and monetise data from other countries without remuneration or adequate regulation.
This refers to the potential for AI leaders to globally hoover data, process it with advanced computing powers, and potentially control other countries’ populations and ecosystems without fair compensation to the proprietors.
Furthermore, AI-driven automation is expected to reduce jobs in terms of numbers, configurations, and intensities. This impact is particularly concerning for countries with booming demographics, such as the Muslim world, sub-Saharan Africa, and non-Fareast Asia, where AI’s job-reducing potential is less suited to their demographic trends.
This could exacerbate existing global inequalities, with a significant proportion of the global population potentially missing out on AI’s positive benefits, especially in poorer countries. The concentration of AI benefits in a limited number of sectors, such as financial services, IT, and healthcare, further contributes to this disparity.
The urgent need for a new social contract to tackle disruptive technologies like AI safely and ethically is paramount, particularly for the Global South, which often experiences low trust rates between its citizenry and government. There are calls for universal, multilateral governance to ensure that AI serves the common good and prevents increased planetary confrontation.
What Comes Next
The overall impact of AI on global affairs is an intensifying challenge that demands urgent attention to establish global governance frameworks and ethical standards for safe and responsible AI deployment. Societies are largely unprepared for the ethical, legal, socio-political, and practical implications of AI deployment. Existing multilateral institutions like the UN and its agencies, designed 75 years before the emergence of these technologies, are poorly equipped to offer comprehensive and timely AI governance.
The speed of technological innovation outpaces any administrative response, and new political trends of “neo-nationalism” further trivialise the capacity of multilateral forums to play a norm-setting and monitoring role.
The choices made today will shape whether AI becomes a catalyst for global progress or a point of contention. Leadership in AI requires more than just technological advancement; it demands the creation of a global framework that promotes collaboration, balances power, and safeguards humanity.
The future international order will be significantly shaped by how the US and China manage their competition and whether they can collectively steer AI toward global progress rather than contention.
Conclusion
The US-China AI Race is undeniably the defining geopolitical contest of the 21st century, fundamentally reshaping global power dynamics and presenting humanity with unprecedented opportunities and profound risks. From the heated battle over semiconductor supply chains and the rapid advancements in data and algorithms, to the fierce competition for talent and the intricate geopolitical ripples of shifting alliances, AI is at the core of every strategic calculation.
The military implications are transformative, promising autonomous weapon systems and enhanced cyber warfare capabilities, yet simultaneously raising the spectre of miscalculation, unintended escalation, and the shared threat of AI misuse by rogue actors. Furthermore, AI’s pervasive influence on disinformation, through sophisticated deepfakes and the amplification of foreign interference, threatens the very foundations of truth and trust in democratic societies.
As we stand at this pivotal crossroads, the future trajectory remains uncertain: will intense competition yield to necessary cooperation to mitigate existential risks, or will it descend into an unpredictable confrontation?
The impact on developing nations, the ethical dilemmas, and the urgent need for robust global AI governance underscore the collective responsibility of the international community.
The rapid pace of AI innovation demands a proactive, multilateral approach, balancing national interests with universal ethical standards to ensure that AI serves the common good and paves the way for a more stable and prosperous future, rather than an era of heightened global instability and inequality. The stakes could not be higher, and the choices made today will irrevocably shape tomorrow.
Post a Comment