Megha Shrivastava is a Doctoral Research Scholar and Dr. TMA Pai Fellow at the Department of Geopolitics and International Relations, Manipal Academy of Higher Education (Institution of Eminence), Manipal, India.
China’s launch of DeepSeek’s R1 model in January sent shockwaves worldwide. Despite US sanctions on advanced hardware, the model’s rapid technological rise, effective performance, and lower production cost compared to the Western rival models, yet again, showcases China’s resilience.
For decades, Silicon Valley has maintained dominance in disruptive innovations, particularly through closed-source, proprietary AI models. However, Beijing’s strategic embrace of open-source technologies signals a shift in the global innovation paradigm. The US-China rivalry now transcends competition for high-end semiconductor access; it represents a deeper ideological and structural divide in AI governance. The emergence of China’s AI model—faster, cost-effective, market-driven, and collaborative—contrasts sharply with the Western approach, which has long prioritized expensive, capital-intensive proprietary models.
While American AI policies and big tech firms have sought to consolidate computational power within a few dominant players, DeepSeek’s success demonstrates an alternative path—one that leverages open-source collaboration to drive efficiency and scalability.
Unlike the West, wherein the big tech companies aim to retain their market dominance through asserting control over expensive proprietary models, China’s open-source innovation model is set to disrupt the game of innovation capability. Not only has DeepSeek refuted the reliance on expensive hardware like Nvidia’s high-end chips, but it also reignites debates on AI dominance, AI sustainability, security, and an emerging world order.
DeepSeek’s rise demonstrates China’s AI development strategy, which is led by open-source collaboration. This strategy is cheaper, faster, more efficient, and most importantly, can bypass the US-dominated restrictive tech infrastructure and resources. Such a strategy is visible in China’s state policy support that aims to reduce reliance on foreign hardware and software.
Following Washington’s tightening of sanctions on Chinese Chip companies like ZTE and Huawei in 2018 and 2019, China witnessed a surge in domestic initiatives to advance open-source AI. Unlike in the West, where open-source projects are largely decentralized and independent, China’s state-led model treats open-source AI as a national strategic asset. The National AI Open Innovation Platform—comprising 23 leading private enterprises—has facilitated access to AI datasets, toolkits, and libraries, fostering an ecosystem of domestic deep-learning frameworks.
Consequently, numerous open-source deep learning platforms have been developed, aimed at scaling domestic computational power. Some notable achievements include Alibaba’s X-Deep Learning (X-DL), Huawei’s MindSpore, and Baidu’s Paddle Paddle, Alibaba’s Qwen among others. The widespread government-backed efforts to foster the development of open-source culture is one of the ways China adapted against the US-led sanctions.
In 2023, China’s state-owned China Electronics Corp. launched its first open-source operating system, called Openkylin, which is based on Linux and developed by a community of 4000 developers. A dozen other Chinese companies are involved in the development of home-grown operating systems that could replace Microsoft’s Windows and Apple’s MacOS operating systems. China’s UnionTech Software Technology Co Ltd. is one of them, developing its “Unity operating system”.
Moreover, China’s AI advancement is reinforced by its centralized governance structure, which facilitates extensive data collection and algorithmic refinement. Unlike democratic regimes where privacy and data regulations impose constraints, China’s authoritative control enables the development of advanced AI systems with vast datasets. China’s open-source driven AI strategy does not remain out of the shadows of China’s sophisticated surveillance system.
The Chinese government is actively promoting open-source code platforms like Gitee, a Chinese rival of GitHub. Unlike GitHub, Gitee is subject to Chinese censorship, reflecting how the Chinese state seeks to align its open-source collaborative approach with its national and strategic priorities. Despite promoting open access provision for independent developers, the platform remains primarily dominated by major companies operating within the strategic framework set by the Party. One such criterion is to demonstrate a consistent commitment to innovate aggressively in order to expose the limitations of Western sanctions. However, this also demonstrates the shortcoming of China’s open-source innovation, in which the government’s tightly controlled big-shots take priority over emerging start-ups.
Another unique and instrumental factor reinforcing the growth in China’s AI industry is the creative alignment with the Communist Party’s strategic goals with its private sector’s commercial objectives. The success of Chinese AI companies demonstrates the touted credibility of the 2020-2022 tech crackdown— the Chinese state’s regulatory spree to strictly align the interest of the private firms with the Party’s objectives.
Instead of raising state-business conflicts, China’s private companies strategically utilised the state’s stringent actions to fix bureaucratic inefficiency and incentivise economic growth, besides supporting the Party’s yet another drive to maintain regime stability. This ultimately strengthened the narrative that state intervention is instrumental in fostering innovation. This integration of state-led directives and private sector innovation has allowed Chinese AI firms to rapidly scale their capabilities while aligning with Communist Party priorities.
The recent measures also suggest a significant departure from the conventional features of China’s industrial policies, which focused on cultivating national champions through direct state sponsorship. Instead, the latest trend suggests a shift towards a more collaborative approach to incentivize a diverse network of AI firms. This network includes the collaboration between state laboratories, academic institutes, the private sector, venture capitalists, etc. This diversified approach aims to accelerate AI commercialization across key sectors, including healthcare, manufacturing, agriculture, and robotics.
Against the West’s corporate-driven approach, China’s renewed strategy demonstrates the emergence of an alternative framework—one that blends open-source innovation with strong state oversight. Beneath all the hypes, the intensifying race between the US and China largely exposes trends that AI development is increasingly driven using economic and political control.
However, as China scales its AI ambitions, it faces a critical challenge—balancing the advantages of open-source collaboration with the risks of a tightly controlled, surveillance-driven innovation environment. The extent to which China can sustain its AI growth while managing these contradictions will determine its long-term success in shaping the global AI order.
Disclaimer: The views expressed in the article are personal.
