A sweeping analysis of almost 14 million Chinese patents reveals that universities, not state-owned enterprises or returning overseas talent, are driving a large share of the country’s recent gains in areas the US labels as critical technologies.
Universities at the frontline of Chinese innovation
The National Bureau of Economic Research (NBER) study examined patenting across 14 technology areas the US Department of Defense considers critical, including advanced computing, artificial intelligence, hypersonics, space technologies and biotechnology. The researchers found that Chinese universities are responsible for more than 25% of inventions in those fields—about eight times the rate attributed to US universities, which account for roughly 3.3%.
That pattern contrasts sharply with the United States, where academic patenting in these sectors has traditionally been modest. In China, by contrast, universities have become a central node in an innovation ecosystem the study describes as “much more diverse—spread out across the parties.”
“We’ve clearly seen this dramatic boom in terms of research funding and patenting in China,” the study’s co-author told reporters, adding the research team had been “programmed” to expect big corporate names to dominate.
State‑owned enterprises and returning talent play smaller roles
The study also undercuts another common assumption: that China’s advances in critical technologies depend principally on state enterprises or the repatriation of overseas researchers. State-owned enterprises and government actors account for only about 4% of the patenting in these fields. Fewer than one in ten of the critical technology patents involve an inventor with US work experience.
Those findings suggest Chinese innovation in strategically important domains is more homegrown and institutionally broad-based than many Western observers have assumed. The researchers said they had anticipated a pattern mirroring historical Western models—large firms driving patenting the way IBM and Samsung once did—but instead found a markedly different distribution of inventors and filers.
- Chinese universities: >25% of patents in 14 critical technology areas
- US universities: ~3.3% in the same fields
- State and government entities in China: ~4%
- Inventors with US experience: <10% of critical technology patents
| Entity | Share of critical tech patents |
|---|---|
| Chinese universities | >25% |
| US universities | ~3.3% |
| Chinese state-owned / government | ~4% |
| Inventors with US work experience | <10% |
Implications for policymakers and industry
The study reframes how competition in advanced technologies should be understood. If universities are principal contributors to patenting in critical fields, then strategies that focus solely on corporate rivals or on stem‑cell repatriation of talent will miss a major source of capability. Policymakers must therefore consider the role of higher education institutions, research funding, and university‑industry linkages when designing responses.
For industry, the distribution of patenting suggests a rich and decentralised pipeline of ideas and prototypes emerging from academic labs, potentially increasing the pool of partners, spin‑outs and collaborators. It also raises questions about how intellectual property is commercialised in China, and how that process interacts with government priorities and industrial planning.
Observers will also see the findings in the context of high‑profile technological moments—such as the release of low‑cost, capable Chinese AI models or unexpected military tests—that have prompted comparisons to historical shocks in US science and technology policy. The study warns against interpreting such events as the product of a few dominant firms alone.
Ultimately, the research offers evidence that the centre of gravity in Chinese innovation is broader and more institutionally distributed than many in the West have appreciated. That carries strategic implications for how the UK and its allies assess risks, target safeguards, and design collaboration or competition policies in areas seen as nationally sensitive.
The study’s detailed dataset and further analysis will be important reading for government advisers and corporate strategists seeking to update assumptions about where and how critical technological advances are being made.
