Image Source: Getty
The emergence of DeepSeek as a formidable Artificial Intelligence (AI) contender last week has raised unsettling questions about the conventional wisdom surrounding AI development—particularly the belief that winning the AI race is purely a function of pouring billions into graphics processing units (GPUs). Despite operating with seemingly fewer and less advanced chips, DeepSeek has managed to produce models that rival America’s best, challenging Nvidia chip company’s dominance in AI infrastructure.
Market signals suggest investors remain steadfast in their faith in the American AI chip giant. Some dismiss DeepSeek’s efficiency claims as posturing, but others see merit. Even if true, it may have simply optimised around American models trained on superior hardware. The stark reality? A Chinese AI now matches the best US models—at a fraction of the cost.
Despite operating with seemingly fewer and less advanced chips, DeepSeek has managed to produce models that rival America’s best, challenging Nvidia chip company’s dominance in AI infrastructure.
Unsettling implications for US
If DeepSeek’s claims prove true, Nvidia’s stranglehold on the AI hardware market could weaken, forcing a rethink in how AI scaling is approached. The real threat to Nvidia isn’t just competition from another chipmaker—it’s the possibility that the industry may soon realise it can achieve cutting-edge AI with far fewer of Nvidia’s products than previously believed. The US$593 billion loss in Nvidia’s market value in one single day is a reflection of those sentiments.
This comes at a time when other American tech companies like Microsoft and Meta are committing vast sums to build GPU-packed data centres, reinforcing the narrative that computational power is the key to AI supremacy. However, this also indicates that DeepSeek’s efficiency signals a potential paradigm shift—one where training and running AI models might not require the exorbitant processing power once assumed necessary.
The DeepkSeek’s emergence also poses a challenge for the Trump administration, which has made AI an immediate priority. On January 21, President Donald Trump unveiled a plan for private sector investments of up to US$500 billion to build AI infrastructure to surpass US competitors in this crucial technology. Emphasising the continued significance of American intellectual capital in maintaining a competitive edge, his administration has pledged to double investments in AI research, created the nation’s first AI research institutes, and introduced the world’s first regulatory guidelines to oversee AI development in the private sector.
President Trump has framed the rise of DeepSeek as both a significant challenge and a catalyst for reinvigorating American AI enterprises.
As expected, President Trump has framed the rise of DeepSeek as both a significant challenge and a catalyst for reinvigorating American AI enterprises. He urged American tech companies to avoid stagnation and reassert their longstanding leadership in technological innovation. Trump’s remarks reveal the critical need for sustained investment in research and development by the American tech ecosystem to ensure continued dominance in an increasingly competitive global landscape.
However, there are also concerns related to Intellectual Property (IP), as suggested by White House AI and cryptocurrency czar David Sacks, who said that DeepSeek may have leaned on the output of OpenAI’s models to help develop its technology. He highlighted “distillation”—a methodological approach wherein one model assimilates information from another—as a potential vector of technological appropriation. If true, this would further strengthen the American establishment’s age-old concerns over Chinese theft of American IP.
Race for AI dominance
AI is not merely another technological advancement; it is the dominant force shaping this decade and likely beyond. Its benefits, like its consequences, follow a deeply non-linear trajectory. The US and China, as the only countries with the scale, capital, and infrastructural superiority to dictate AI’s future, are engaged in a race of unprecedented proportions, pouring vast sums into both model development and the data centres required to sustain them.
For the longest time, Washington operated under the assumption that it was unassailably ahead of China in AI and was determined to keep it that way by restricting the necessary tech to China. The thinking was simple: cut off China’s access to critical hardware, and its AI progress slows down. The Biden administration, for instance, doubled down on restrictive measures—banning the export of advanced chips and AI-related tech to Chinese tech companies. However, DeepSeek’s success suggests that the US approach may have yielded unintended consequences. By forcing Chinese companies to get scrappy and optimise every last bit of their available limited computing power, the US may have made them more efficient. This is a classic case of second-order effects. The US thought it was setting up guardrails, but instead, it created the kind of adversity that breeds innovation. The question now isn’t whether China can catch up—it’s whether the US can move fast enough to stay ahead.
This positioning is a direct challenge to America’s technological dominance, underscoring China’s growing capabilities and ambitions to carve out a parallel tech empire.
Besides, Beijing has been methodically executing a long-term strategy to establish its domestic tech capabilities, and the rise of DeepSeek represents another key milestone in this pursuit. The timing and messaging surrounding the development of this AI technology seem strategically designed to send a clear signal to the world: at a time when President Trump is considering more tariffs and restrictions, China would like to claim that US export controls are not as effective as intended, and the era of America’s undisputed leadership in AI may be coming to an end. This positioning is a direct challenge to America’s technological dominance, underscoring China’s growing capabilities and ambitions to carve out a parallel tech empire.
While China may temporarily achieve cost-efficiencies in AI production, long-term success in this field will ultimately depend on foundational innovation—a strength historically held by the US. This perspective aligns with economic theories that suggest initial innovation and creative capabilities are critical drivers of sustained competitive advantage, even in the face of shifting global dynamics and technological advancements. The underlying message is that while short-term efficiencies can be replicated, lasting dominance is rooted in original intellectual contribution.
Conclusion
The emergence of DeepSeek may or may not be a Sputnik Moment of the 21st century. Yet, the reality is that as of early 2025, a Chinese AI now matches the best models from the US—at a fraction of the cost. Silicon Valley has had its awakening: there are now more cost-efficient and faster ways to develop AI, and it’s no longer just the American way. Washington, too, has come to realise that China is fully up to the challenge, even as the rest of the world is savouring the fruits of these technological advancements.
As the US deliberates its forthcoming policy interventions, the regulatory and strategic response to DeepSeek will inevitably shape the broader contours of AI governance, intellectual property jurisprudence, and international technological sovereignty. The discourse, much like the field it pertains to, remains a dynamic interplay of risk, uncertainty, and competitive asymmetry.
As the US deliberates its forthcoming policy interventions, the regulatory and strategic response to DeepSeek will inevitably shape the broader contours of AI governance, intellectual property jurisprudence, and international technological sovereignty.
This AI race is no longer a sprint but a marathon. Runners lead at different points, but as the barriers to entry lower, new competitors are emerging—ones who might not have had the capital to play the game before. The next few years will be decisive, as the real game will not just be about who leads but about who understands how to thrive in a system that’s becoming increasingly unpredictable.
Moreover, the real impact of this race lies in the second-order effects—on productivity, economic asymmetries, and systemic fragilities that are neither immediately visible nor easily quantifiable. The real question isn’t who’s ahead in AI but how the unintended consequences—power shifts, efficiency gains, and hidden risks—ripple through an already fragile and polarised geopolitical landscape.
Sameer Patil is the Director of the Centre for Security, Strategy and Technology at the Observer Research Foundation.
Sauradeep Bag is an Associate Fellow with the Centre for Security, Strategy, and Technology at the Observer Research Foundation.
The views expressed above belong to the author(s). ORF research and analyses now available on Telegram! Click here to access our curated content — blogs, longforms and interviews.
PREV
