DeepSeek’s rapid rise to fame this month after the release of its new AI chatbot has supercharged debates about the efficacy of U.S. export controls and the race to AI-tech superiority. Many in China celebrated DeepSeek’s open-source AI model—ostensibly built with a fraction of the resources invested into the closed-source leading American counterparts, and performing just as well—as a national victory. DeepSeek displaced ChatGPT as the U.S. App Store’s top app, after RedNote took the same crown earlier this month. China Daily published a triumphant op-ed about the company’s AI model titled, “R1 exemplifies [the] open spirit of the internet.” Zhou Hongyi, co-founder, chairman and chief executive of Chinese cybersecurity firm Qihoo 360, declared that DeepSeek has “upended the world” and that “we should have confidence that China will eventually win the AI war with the U.S.” As CDT Chinese editors documented, users on Zhihu have expressed interest, pride, and even skepticism about what DeepSeek AI and other recent technological advances mean for China’s future:
Zhihu question: "What do you think of video game developer Feng Ji [CEO of Game Science and creator of the video game ‘Black Myth: Wukong’] calling DeepSeek a "scientific achievement that could alter China’s destiny"?
Answer from Zhihu user Zhang Shan:
China’s sixth-generation fighter jets are our "national destiny," as is Xiaohongshu, as is DeepSeek. And 2025 has only just begun—is this what’s in store?
What’s our next "national destiny"?
We’re "winning" so much it’s scary. Things haven’t been this exciting for a century. [Chinese]
Western media has also picked up on a sudden shift in the global AI tech race, or at least the perception of it. Numerous outlets have described DeepSeek’s achievement as a potential “Sputnik moment,” and a recent Economist piece was titled, “China’s AI industry has almost caught up with America’s.” MERICS senior analyst Antonia Hmaidi stated that it is “something that has really shown to some degree how much the U.S. was living in a bubble.” Monday’s Wall Street Journal editorial headline read, “The DeepSeek AI Freakout: The Chinese startup’s model stuns Big Tech—and Wall Street—with its capability and cost.” The release of DeepSeek’s R1 chatbot touched off a market rout which wiped out $1 trillion in value from the NASDAQ 100, including almost $600 billion from AI chipmaker Nvidia. AI reporter Karen Hao argued in a thread on BlueSky that DeepSeek’s success represents a big crack in the current dominant paradigm of AI development, undercutting the main logic behind the U.S. government’s astronomical Stargate AI investment pledge and highlighting the negative externalities of inefficient AI development that prioritize financial growth and monetization over science. At Foreign Policy last week, Howard French argued that the Chinese government’s commanding approach to big tech, in contrast with that of the U.S., has helped to prevent greedy tech oligopolies from hijacking the country’s political system and fiscal resources:
In bringing this new class of business titans to heel, China’s leaders made a carefully considered strategic decision about the direction of their country’s political economy. In effect, they were saying that Beijing would never grant a dominant role to the extraordinarily lucrative and freewheeling private technology sector. Put slightly differently, that sector would have no sacred cows and would never be allowed to cast a shadow on the party and state.
[…] China has not reined in its tech sector out of any belief in democracy, but rather through a seeming understanding that the new forces of wealth, data, intelligence, information, commerce, and communications can hijack a country’s political system and lead it into dangerously uncharted territory.
Trump, who betrays little technical sophistication, has done the opposite, as he has embraced the big tech sector and celebrated its wealthiest. If this is not challenged, the world might one day look back at this time as the moment when the U.S. state was captured. [Source]
Caught off guard by the unexpected turn of events, “American AI firms [tried] to poke holes in disruptive DeepSeek,” as a Reuters headline read, suggesting that the publicly available figure for the cost of its training “is the tip of the iceberg in terms of total cost,” and that its vaunted efficiency is therefore exaggerated. Many articles discussed to what extent DeepSeek may have “outsmarted America” by exploiting “loopholes in existing export controls.” AI researcher Matt Sheehan explained, “The US export control has essentially backed Chinese companies into a corner where they have to be far more efficient with their limited computing resources.” Commenting on the talent of DeepSeek’s young staff, Marina Zhang, an associate professor at the University of Technology Sydney who studies Chinese innovations, told WIRED, “This younger generation also embodies a sense of patriotism, particularly as they navigate US restrictions and choke points in critical hardware and software technologies. Their determination to overcome these barriers reflects not only personal ambition but also a broader commitment to advancing China’s position as a global innovation leader.”
However, writing in his Value Added Substack about the conditions that led to DeepSeek’s success despite all of these impediments, JS Tan argued that DeepSeek is actually an outlier in China’s AI innovation ecosystem:
I’ve seen many recent takes that frame DeepSeek as evidence that China’s innovation system is thriving. If we count any innovation that originates from a Chinese firm as part of China’s innovation system, then I suppose this is true.
However, I think this framing is a misreading of the situation. Unlike the rise of China’s electric vehicles, batteries, or even solar PV, DeepSeek is neither a state-led project nor a direct beneficiary of China’s AI-focused industrial policies. In this way, it’s an outlier, not a representative case, within China’s innovation ecosystem. The fact that the company was self-funded by a former hedge fund manager and emerged from the periphery of China’s tech landscape rather than from the country’s established AI giants like Baidu or Tencent is a testament to this point.
Asking how China’s innovation system produced DeepSeek is, therefore, the wrong question. DeepSeek’s success arose not because of China’s innovation system but, in many ways, in spite of it. Here, I’ll highlight three important ways [human capital, funding and corporate governance structure, and labor practices] in which DeepSeek diverges from the innovation model of China’s tech sector. [Source]
Other analysts debated the efficacy and relevance of export controls on China’s AI advances and how DeepSeek’s breakthrough will affect future U.S. strategy. Lennart Heim and Sihao Huang wrote in Heims’s AI governance blog about the ways in which the coverage of DeepSeek’s achievements overlooks the complexity of (and U.S. strategic advantages in) compute access, export controls, and AI development. Dean W. Ball argued at Lawfare that DeepSeek’s AI breakthrough does not invalidate the U.S.’ basic export control strategy, since the latter is designed more to deny China an AI computing ecosystem rather than prevent the emergence of a single, high-performing Chinese model. Paul Triolo suggested in his AI Stack Decrypted Substack that the U.S. and China should increase collaboration on AI safety now that advanced AI technology is more easily available to malicious actors, and he challenged the common assumption that attaining nuanced U.S. export controls is even feasible, let alone cost-effective. Lizzi C. Lee argued at The Wire China that the current U.S. approach to securing leadership in AI by erecting barriers is fundamentally flawed:
What makes DeepSeek’s story significant — and what should give U.S. policymakers pause — is its ability to thrive despite the constraints of sanctions. Denied access to advanced semiconductors like Nvidia’s H100 GPUs, its lab adapted by developing bespoke algorithms that optimized less sophisticated H800 chips. These innovations have mitigated the hardware limitations placed on DeepSeek, offering a glimpse into a deeper truth: constraints can be powerful catalysts for creativity. When faced with scarcity, talented researchers often devise solutions that not only overcome immediate hurdles but also position them for longer-term breakthroughs. DeepSeek’s success illustrates how sanctions, while disruptive, can inadvertently fuel the innovation they seek to stifle.
The impact of these sanctions extends well beyond DeepSeek’s lab. Across China, U.S. export controls have spurred an accelerated push toward technological self-sufficiency. Beijing has committed billions to building a domestic semiconductor industry, pouring resources into research, infrastructure, and workforce development. While China remains years away from achieving parity with the United States in advanced chip manufacturing, its trajectory suggests that the technological gap is narrowing. The more the United States focuses on containment without addressing its own vulnerabilities, the more it risks hastening the erosion of its lead.
[…] Dismissing DeepSeek’s success as an isolated fluke would be shortsighted. Its achievements illuminate both the ingenuity born of constraints and the cracks in America’s current strategy. More than a technological feat, DeepSeek signals the urgent need for a broader vision. [Source]
Translations by Cindy Carter.
