At a high-tech fair in Shenzhen last month, surrounded by humanoid robots performing everything from kick-boxing to piano playing, a question echoed among tech influencers that would have seemed absurd two decades ago: Can the West catch up with China? According to the latest data, this isn’t just speculative chatter�it’s a reality check for global AI leadership?
The Research Gap Widens
The Australian Strategic Policy Institute’s critical technology tracker reveals China now leads in 66 of 74 high-impact research areas, including computer vision and quantum sensors, while the US leads in just eight? This represents a dramatic reversal from 2005, when China accounted for only 6% of the world’s most highly cited research papers compared to 43% for the US? Today, China’s share has surged to 48%, while America’s has plummeted to 9%?
“China is building the whole technology ecosystem,” says Jenny Wong-Leung, one of the report’s authors, highlighting how persistent long-term investment in fundamental research has paid off while the US defunds many federal science programs? Nature’s latest ranking of research institutions confirms this shift, with nine of the world’s top 10 research institutions now Chinese, leaving only Harvard University in the top tier?
Two Competing AI Philosophies
The most consequential battleground is artificial intelligence, where the US and China are pursuing fundamentally different approaches? American tech giants like OpenAI, Alphabet, Microsoft, Meta, and Amazon are making colossal investments in massive, proprietary “closed-weights” models such as ChatGPT and Gemini, designed to achieve generalizable intelligence? OpenAI alone plans to invest $400 billion over the next few years to build out its Stargate data centers across the US?
Meanwhile, Chinese companies favor smaller, cheaper “open-weights” models like DeepSeek and Alibaba’s Qwen that can be more readily adapted by developers? This approach partly reflects necessity�US export restrictions have denied China access to state-of-the-art silicon chips needed for the most powerful foundation models�but also represents a strategic choice to rapidly diffuse the technology throughout the economy?
The Sustainability Question
IBM CEO Arvind Krishna offers a sobering perspective on the US approach, criticizing the current AI infrastructure spending race as economically unsustainable? “It costs about $80 billion to build a data center with a capacity of one gigawatt,” Krishna notes, adding that scaling to 100 gigawatts would require $8 trillion in investment, making profitability impossible? He also expresses skepticism about achieving Artificial General Intelligence with current technologies, estimating only a 0-1% chance?
This skepticism finds support in recent market developments? Microsoft has lowered sales growth targets for its AI agent products after many salespeople missed their quotas, with enterprise customers resisting premium prices for tools prone to confabulation and errors in novel scenarios? The company reported capital expenditures of $34?9 billion for the fiscal first quarter ending October 2025, yet less than a fifth of salespeople in one US Azure unit met 50% growth targets for their Foundry product?
Open Models Gain Ground
A recent study by MIT and Hugging Face found Chinese open models have now overtaken comparable US models in global adoption? Many US companies, including Airbnb, have become fans of the “fast and cheap” Qwen model? Michael Power, former global strategist at investment firm Ninety One, argues the US is making a “catastrophic strategic error” in betting so heavily on giant closed AI models?
“China’s model is turning out to be far more effective in terms of usable compute in the real world,” Power explains, especially considering the country’s lower energy costs? Even Sam Altman, OpenAI’s chief executive, has expressed concern that “we have been on the wrong side of history here?”
The Regulatory Landscape
While technological approaches diverge, regulatory battles intensify domestically? A Republican-led effort to include a measure blocking state AI laws in the National Defense Authorization Act has failed due to bipartisan opposition? The measure, backed by President Donald Trump, aimed to prevent states from passing AI regulations for a decade, arguing that a patchwork of state laws would hinder innovation and allow China to catch up?
House Majority Leader Steve Scalise argued, “We MUST have one Federal Standard instead of a patchwork of 50 State Regulatory Regimes? If we don’t, then China will easily catch us in the AI race?” However, opposition came from both Republicans like Marjorie Taylor Greene and bipartisan groups like Americans for Responsible Innovation, which advocates for state-level AI safety laws?
Broader Implications
The debate extends beyond technology to economic fundamentals? While Anthropic, an AI company rivaling OpenAI, prepares for an IPO that could value it at $350 billion next year�when it would be just five years old�questions remain about sustainable business models? The company projects $70 billion in sales by 2028 but hasn’t yet achieved profitability, highlighting the gap between valuation and actual financial performance in the AI sector?
Meanwhile, in healthcare, AI’s impact proves more complex than early predictions suggested? Despite Geoffrey Hinton’s 2016 prediction that AI would outperform radiologists within 5-10 years, radiologist numbers have actually increased by over 40% in the UK NHS since 2016? AI tools are used alongside rather than replacing radiologists, creating new tasks like post-deployment monitoring and speeding up workflows that lead to more images to analyze?
A Strategic Crossroads
The fundamental question isn’t just about who leads in research papers or who builds the largest models, but which approach creates more sustainable value? China’s strategy of open, adaptable models diffused throughout its economy contrasts sharply with America’s focus on massive, centralized systems requiring unprecedented infrastructure investment?
As David Lin, a senior adviser to the Special Competitive Studies Project, observes, “We’re essentially pitting our private capitalists against this nation state of China? The stakeholders here have two very different sets of resources, attributes, strengths and weaknesses?” This asymmetry in approaches may ultimately determine not just who wins the AI race, but what kind of AI future emerges for the global economy?