The landscape of artificial intelligence (AI) is rapidly evolving, particularly with the emergence of Chinese companies like DeepSeek and Moonshot AI, which are creating waves in the global market.
In early 2025, the release of DeepSeek’s R1 artificial intelligence model led to rising concerns within U.S. policy circles, as it competes directly with advanced proprietary AI models from the U.S. Despite ongoing American export controls on sophisticated semiconductors, DeepSeek has managed to pioneer an open technology that poses a significant threat to U.S. leadership in AI.
The stakes escalated further with the introduction of Kimi K2, an autonomous AI model capable of complex tasks by Moonshot AI, prompting discussions about another ‘DeepSeek moment’ in international AI dynamics.
However, the crux of the threat from these Chinese open models is not just about competition in AI capabilities.
It extends to the broader implications of AI adoption worldwide, particularly in developing countries eager to harness AI’s potential in various sectors.
As of January 2025, the DeepSeek R1 app reported 33 million active users globally, a number that surged to nearly 97 million by April.
Additionally, over 500 customized derivative versions of the R1 model had been downloaded a total of 2.5 million times within the same month, indicating a strong demand for adaptable AI solutions.
This trend highlights the effectiveness of China’s low-cost, open-model approach, providing a distinct advantage in meeting the needs of researchers and developers, particularly in developing regions.
The implications of which country’s AI models dominate the global sphere reach beyond mere economic competition and military applications.
Open models like R1 and Kimi K2 provide opportunities for countries around the world to customize AI systems for local needs, potentially in healthcare, education, and workforce development, all at lower costs than American counterparts.
In this light, the soft power benefits that Chinese open models could offer might reshape international perceptions, positioning China as a benefactor of AI technology in diverse regions including Africa, Asia, Latin America, and the Middle East.
By sharing the benefits of open AI, Chinese models could garner goodwill on the world stage, leveraging AI’s potential for economic and social development.
In contrast, the United States, where companies primarily focus on closed AI models, faces implications tied to its security-focused policies that minimize technology sharing due to risks of adaptation for military or economic purposes by rival nations like China.
America’s AI strategy may indeed be risking the loss of influence and strategic leverage in essential regions of the world if it fails to address the rise of open models.
As policymakers consider the road ahead for AI, the need for the U.S. to balance national security risks with the promotion of innovative technology to the global community has never been more pronounced.
China’s soft power strategy in AI could be fortified by its current limitations in accessing computing power, creating an opportune moment for the U.S. to adjust its tactics.
Of particular note, the unique approach of DeepSeek lends itself to a model of smaller, more efficient AI solutions that are economically viable for many markets, diverging from the traditional focus on building the most powerful general-purpose models.
This adjustment stems from a practical need brought about by U.S.-led export controls and targets a fulfillment of real-world applications rather than purely theoretical advancements.
China’s open models closely resemble the capabilities of leading closed models from U.S. companies, raising the possibility of broader adoption as countries look for foundational AI tools.
Beijing’s strategic objective of expanding soft power, particularly in the global South, is built around sharing advanced technologies to facilitate domestic innovation and reduce reliance on Western tech.
In February, a Chinese Foreign Ministry representative underscored the intention to disseminate the advantages of open AI technologies to international partners.
Within the framework of international relations theory, soft power acts as a mechanism for influencing other countries through shared values and ideas, contrasting with hard power, which relies on military dominance.
Historically, the United States has successfully harnessed its cultural appeal and economic strength to forge strong alliances, leading to significant technological advancements.
However, the dual-use nature of AI complicates its categorization within soft power strategies, as both nations vie for hard-power applications in military contexts with autonomous drones and advanced command systems.
Growing concerns regarding China’s potential military use of AI technologies have prompted U.S. policies that lean toward restricting access to critical hardware necessary for the development of advanced AI models.
Yet, this narrow focus neglects the transformative potential of AI in driving global economic growth and enhancing societal well-being, particularly in under-resourced areas.
If American AI models can be harnessed to develop new treatments for diseases or improve agricultural practices in developing countries, this would enhance U.S. diplomatic standing and exemplify the benefits of American technology.
China, on the other hand, is strategically focused on distributing its technologies while simultaneously improving its image as a global supplier of cutting-edge AI solutions.
The difference between open and closed model ecosystems is crucial in determining each nation’s approach to the global AI landscape.
Open models, like DeepSeek’s R1, feature open weights that allow developers to adapt and optimize them for particular activities, while closed models – such as those from OpenAI – remain inaccessible beyond tightly controlled interfaces.
Though closed models tend to be more powerful, their restrictions have made them less appealing for global AI developers compared to the flexibility offered by open models.
Recent innovations, notably by DeepSeek and Moonshot AI, indicate that the performance gap between these two paradigms is closing, accelerating the potential for widespread adoption of Chinese technology.
China has also intensified its strategy to promote soft power as it develops a robust digital infrastructure through initiatives like the Digital Silk Road.
Chinese tech giants such as Huawei and Alibaba Cloud are expanding their global presence by investing heavily in telecommunications and cloud computing infrastructure across developing nations.
These investments create a solid foundation for deploying AI applications around the globe, especially if China can circumvent U.S. restrictions on high-end chip access in the future.
In pursuit of its strategic objectives, China is not only enhancing its technological infrastructure but is also working to solidify control over the global narrative surrounding AI.
This includes initiatives such as the Global AI Governance Initiative, which aims to foster responsible AI development while positioning China as a key player in guiding global policy.
U.S. policymakers, however, need to carefully reassess their strategies in light of the advancements made by Chinese companies, recognizing the inherent risks in prioritizing closed models.
In response to the growing influence of Chinese open models, major U.S. firms are beginning to rethink their strategies.
Sam Altman, CEO of OpenAI, acknowledged the need for a balanced approach, leading the company to develop an open-source model expected for release in late summer 2025.
Similarly, Google has introduced a family of open-source models, although these still lag behind their closed counterparts in terms of performance.
Recognizing the shifting dynamics, OpenAI recently took significant steps by reducing the pricing of its models, making them more attractive for developers, researchers, and startups to foster competitive parity with Chinese offerings.
The challenge facing the U.S. is that China’s efforts to become the dominant player in AI may face hurdles due to American-led export controls limiting access to the tools necessary for creating advanced chips.
Despite successful initial developments, demand for cutting-edge chips will likely continue to outstrip supply, complicating the ability for Chinese models to achieve mainstream integration.
American policymakers must act with urgency in this context, prioritizing the promotion of an open-model AI ecosystem to reclaim global influence in the face of Chinese advancements.
The AI Action Plan released by the Trump administration acknowledges this strategic need for the U.S., emphasizing the importance of leading open models that reflect American values.
While Meta’s advancements in open modeling are noteworthy, the contributions from various stakeholders in the U.S. landscape must be harnessed to boost the overall ecosystem.
Collaboration between developers and research institutions will be crucial, as will support from entities like the National Artificial Intelligence Research Resource to empower under-resourced developers competing against their larger counterparts.
Revising export control strategies will also be necessary, with an emphasis on developing clear guidelines for chip exports that align with contemporary AI applications.
This recalibration should focus on controlling semiconductors that are pivotal for model inference, presenting a better alignment for current AI market trends.
Enhancing the capability to share the benefits of AI technologies through open-model development can play an important role in advancing America’s soft power across the globe.
Doing so would provide allies and partners with appealing and cost-effective alternatives to Chinese models while simultaneously helping to foster innovation abroad.
Ultimately, if the U.S. does not act swiftly to counter China’s growing influence in soft power through AI deployment, it may find itself at a significant strategic disadvantage in global technology diplomacy.
The potential ramifications of failing to address this situation could be profound, as the diffusion of affordable, powerful AI technologies could elevate China’s position on the world stage while challenging U.S. preeminence.
image source from：foreignaffairs
