AI Patent Activity in Traditional Medicine Is Led by China, but Data and Safety Challenges Remain
A new report highlights China's dominance in patent activity while warning that data quality, clinical reliability and governance remain major challenges.
China accounts for the majority of global patent activity involving artificial intelligence applications in traditional medicine, according to a new industry report presented at the World Conference on Traditional Medicine in Beijing. The findings highlight growing efforts to combine digital technology with traditional medical knowledge, while underscoring continuing challenges involving data quality, clinical reliability and governance.
The report found that 3,030 patents related to AI applications in traditional medicine had been published worldwide since 2000. China accounted for 80.4% of published patents and 84.2% of granted patents, according to Xinhua's report on the findings.
The figures point to China's leading position in this particular patent category, but patent activity alone does not establish that a technology is clinically effective or widely used. The report also emphasized that AI-enabled traditional medicine has moved beyond the conceptual stage, with applications increasingly extending from isolated tools into more coordinated clinical and research workflows.
In traditional Chinese medicine, diagnosis commonly involves observation, listening and smelling, inquiry, and pulse-taking. These methods rely partly on practitioners' experience and interpretation, creating a challenge for developers seeking to translate diagnostic information into standardized data that AI systems can process.
Researchers are exploring tools that can quantify features such as tongue appearance and pulse patterns, helping practitioners record information in more consistent formats. AI may also support the organization of traditional medical knowledge, the analysis of research data and the screening of candidate compounds for further investigation.
The potential applications extend beyond major hospitals. Digital diagnostic tools, remote consultations and AI-supported guidance could help healthcare providers in less well-served communities access information and specialist support. In principle, digital systems could make certain services easier to deliver across geographic distances, although their usefulness will depend on the quality of the underlying data, clinical oversight and the needs of local healthcare systems.
The report also discussed the preservation of traditional medical texts. Digital libraries can make historical material easier to search and organize, while structured databases and knowledge graphs may help researchers identify relationships between concepts, treatments and historical records. Such systems could support education and research, but the interpretation of traditional terminology remains a technical challenge.
The World Health Organization has identified digital and intelligent technologies as relevant to the safe, effective and sustainable development of traditional medicine, Xinhua reported. The wider international interest in digital health gives researchers an opportunity to explore how traditional knowledge can be documented and evaluated while maintaining appropriate standards of care.
Significant obstacles remain, however. The report identified a gap between experience-based terminology in traditional medicine and the standardized, quantitative data required by AI systems. It also called attention to the need to improve model reliability and explainability. If a system produces a recommendation, clinicians need to understand its limitations and determine whether the result is appropriate for an individual patient.
Data protection is another concern. Digitized traditional knowledge can be valuable for research and commercial development, creating questions about ownership, authorized access and the protection of sensitive information. The use of large language models for translation introduces additional risks: a system may produce fluent text while failing to preserve medical meaning or cultural context.
Human review remains essential. Xinhua quoted a healthcare executive who said medical documents at her hospital must be reviewed by physicians to mitigate risks from automated translation. This illustrates a broader principle for AI-assisted healthcare: automated systems can support professional work, but should not replace clinical judgment without evidence and suitable safeguards.
The report called for more consistent data-coding standards, interdisciplinary expertise and international governance mechanisms that balance data security with knowledge sharing. China's patent leadership suggests a substantial base of research activity, but the next stage will depend on whether developers can turn technological progress into trustworthy, transparent and appropriately evaluated applications.