According to the World Intellectual Property Organization (WIPO) Patent Landscape Report of July 2024, China is leading the world in terms of patent applications for generative AI technologies. This advancement is puzzling: Has the global race for AI supremacy altered significantly, or is it just a product of China’s obsession of obsessively stuffing patent offices for years with vaporous documents?
The importance of China’s advantage in generative AI patents will benefit mark China’s shift from being an imitation economy to a true innovator, so does unlocking at least some of the complex value chains which emerging technologies and systems based on self-dependent machine learning-not neural network technologies offer. The four majors have been highly active over the last couple of decades in patenting almost everything that moves or operates within the space of deep reasoning-based artificial intelligence.
China has submitted the most patent applications since 2011 and has been the lead global filer under WIPO’s Patent Cooperation Treaty (PCT) after 2019. However, having the highest number of applications does not correlate to possessing the most innovation and advanced technology. China has always focused on sheer volume at the expense of quality, which is resulting in a plethora of filings without any remarkable advancement in technology. This approach driven by policy has been internalized into the institutional structure of the nation, causing a high number of filings without revolutionary technological advancement.
Shifting Trends in Generative AI Patents
As is notable from history, there seems to be scope for improvement in generative AI patents from China. Companies in the private sector like Tencent, Baidu, Alibaba, ByteDance, and Huawei have dominated the surge in patent applications, especially in subdomains like natural language processing, computer vision, and AI model building.
Moreover, key metrics related to patent quality are also improving, such as grant approval rates and the subsequent success of the commercialization. The approval rate of patents from China in 2023 reached 55% which is a massive increase from 30% in 2019 and has been steadily increasing over four years. In the same period, the rate of commercialization among high-tech firms also increased, suggesting that these patents are now more useful than other patents in other industries.
One of the significant contributors to this positive change is that Generative AI is a young field where China’s leading tech companies and even research institutions such as the Chinese Academy of Sciences, Tsinghua University, and Zhejiang University have significant inputs. The focus on innovation through research rather than through simple political incentivized patent harvesting may be what is improving the quality of these applications.
Challenges Remain in Progress: Focusing on China
Although China’s generative AI patents are likely of better quality than its overall patent portfolio, there are challenges that the country still faces. China continues to perform poorly in grant ratios and international coverage. The Technical Analysis Report on AI Patents from China’s Ministry of Industry and Information Technology for the year 2024 states that the grant rate of China’s next generation AI patents is 32%, with significant differences among big players such as Baidu (45%), Huawei (24%), Tencent (30%), Alibaba (25%), and Ping An (22%).
In addition, China still holds a disproportionate number of patents domestically which severely limits their impact. The percentage of patents that Chinese firms filed outside the country was only 7.3% in 2024, which, while being an increase from 6.3% in 2019, is still extremely low. This stands in stark contrast to developed countries such as Canada (77%), France (65%), Italy (87%), and Japan (70%), which have a much higher filing rate of international patents. Since international patents are generally more scrutinized and have a higher claim to fame, this absence of presence greatly reduces China’s potential to impact the global AI ecosystem.
Effects on the Competition for AI Systems Globally
It remains to be seen whether China’s large volume of generative AI patents will significantly impact the global AI market and will significantly rely on downstream industry developments. At present, generative AI—epitomized by large language models (LLMs) like OpenAI’s ChatGPT and foundational models like Sora—is often viewed as a precursor to artificial general intelligence (AGI). OpenAI, Google, Meta, and Microsoft are leading the world in AGI and AI-powered productivity transformation.
Nonetheless, China still has fundamental obstacles to overcome if it wants to achieve technological supremacy. China has a monopoly on the number of patents filed, but she is behind on three major components of AI innovation; these include sufficient computational resources, availability of high-quality training data, and superior AI chips. The Sanctions imposed by the United States, which limits China’s access to high end semiconductors like NVIDIA’s H100 graphics processing unit, worsens the situation.
Such a technological gap restricts the impact of China’s patents in the region to a theoretical level. Cambricon, a Chinese AI startup with many AI processor design patents, serves as a case in point. The company has been unable to compete globally because they cannot access the advanced chip fabrication technologies needed to create efficient AI hardware.
Generative AI patents are equally incorporated into the core business strategies of major Chinese companies like Ant Group, Alibaba, Ping An, Huawei, Tencent, and even Huawei. While Tencent deploys AI advancements into WeChat Work, Tencent Docs, and Tencent Meeting, Huawei and Oppo switch gears on the application and use AI in telecommunications. AI is integrated by financial sector participants such as Ant Group and Ping An for fintech services. They cannot keep up with the innovation tempo of their American colleagues, however, due to constraints in the quality of data and the availability of chips.
Substance Over Quantity
Although US-based companies have filed fewer patents than China, they still lead in technology innovations. As noted in WIPO’s Patent Landscape Report, US firms’ patents and research citations are considerably higher, demonstrating their more significant impact and recognition across the world. For example, OpenAI, which has published only 48 research papers to date, has been cited an astounding 11,816 times, which clearly illustrates the impact of its contributions.
What stands out in this juxtaposition is a critical fact about Chinese patents: They are showing improvements in quality, but their overall impact is lacking. A dominant position in AI is not defined merely by the sheer volume of patents. It requires essential components such as proprietary research, successful commercialization, and access to important computing facilities alongside the multitude of patents.
In conclusion, China’s rise in generative AI patents is remarkable, but at this point, it is not a game-changer. The emphasis on quantity continues to hamper China’s ability to make a broader impact on the industry. With the persistent limitations regarding technological infrastructure, global patent legitimacy, and access to upper-scale semiconductors, China’s generative AI patents will continue playing a minimal role in determining the future of AI.
