New AI for China's 5 Million Doctors: Exclusive Partnership with Top Journals, Focused on Evidence Sources

Alibaba Health has officially launched a new medical AI product named "Hydrogen Ion" (氢离子), designed specifically to assist China's 5 million doctors in accessing and verifying medical evidence efficiently. The announcement was made at a conference attended by prominent figures from the medical field, including representatives from Peking University, Tsinghua University, and the British Medical Journal (BMJ) Group.

Traditional general-purpose large language models often produce plausible-sounding but incorrect information, a critical flaw in the low-tolerance medical domain. Hydrogen Ion aims to solve this by grounding every answer in verifiable evidence. The AI can answer natural language queries, such as "Latest guidelines on SGLT2 inhibitors for diabetic kidney disease," and provides responses with citations that link directly to the original source text, whether it be clinical guidelines, research papers, or drug instructions.

One of the standout features is the exclusive content partnership with the BMJ Group, which grants Hydrogen Ion access to the full text of 70 medical journals published over the past decade. This makes it the only medical AI assistant in China that allows in-platform reading of BMJ's extensive library. Additionally, Hydrogen Ion has integrated data from domestic authoritative bodies like the Chinese Medical Association and the People's Medical Publishing House.

The AI also addresses the challenges of reading scientific papers. A typical clinical research article can be dissected in 3-5 minutes, compared to the 1-2 hours it might take manually. Furthermore, it provides medical terminology translation and side-by-side Chinese-English reading, as over 80% of doctors in interviews reported needing translation tools for English medical content.

Hydrogen Ion's reliability stems from its four-layer evidence-based AI architecture. The first layer, evidence understanding, uses the PICO framework and GRADE standards to structure medical literature. The second layer performs precise retrieval, matching doctor queries to relevant evidence using semantic understanding beyond simple keywords. The third layer involves fine-tuning the model to emphasize accuracy, fidelity to evidence, and safety, ensuring the AI can clearly communicate when evidence is insufficient or contradictory. The fourth layer is an expert review system, with a medical AI expert committee comprising over 300 Chinese clinicians who continuously validate and provide feedback on AI responses.

The product positions itself as an assistant that does not replace doctors' judgment but rather organizes and presents the best available evidence. In a panel discussion, Dr. Liu Jing from Peking University People's Hospital emphasized the importance of not blindly trusting either outdated guidelines or AI suggestions, but using AI as a bridge to fill information gaps. This approach aligns with the goal of democratizing access to top-tier medical evidence, potentially reducing information disparities between doctors in major hospitals and those in resource-limited settings.

While the long-term impact remains to be seen, Hydrogen Ion represents a shift in medical AI from simply answering questions to providing verifiable, evidence-based responses. The competition in medical AI is no longer about who can generate the most fluent text, but about ensuring every claim has a clear, traceable source.