- China’s AI assistant links medical answers to official drug data, clinical guidelines and research papers
- The move highlights a global challenge for healthcare AI: making machine-generated advice verifiable
Alibaba Group’s AI assistant Qwen has upgraded its health consultation service by adding traceable sources to every medical recommendation, allowing users to see the clinical guidelines, research papers or official drug information behind an answer.
The Qwen app on July 8 integrated five major medical knowledge repositories, including the full database of drug instructions from China’s National Medical Products Administration, a knowledge base from the Chinese Nutrition Society, the latest clinical guidelines, tens of millions of medical papers and more than 10,000 medical publications.
The company said the upgrade gives every health response an “ingredient list” — similar to the nutrition labels on food packaging — showing users where AI-generated advice comes from and what evidence supports it.
AI-powered health consultations have surged in popularity worldwide, but they face a fundamental trust problem: users often cannot tell whether a seemingly professional answer is based on reliable medical evidence or unreliable online information.
Large language models are trained on massive datasets collected from the internet, where accuracy and authority can vary widely.
Traceable sources
In healthcare, where incorrect recommendations can carry serious risks, the lack of traceable sources has become one of the biggest obstacles to wider adoption.
Qwen’s latest upgrade aims to address that issue by transforming AI health responses from a “black box” into a more transparent system.
The platform has established a curated list of medical sources, prioritizing information published by authoritative institutions and leading hospitals.
Its answer-generation process follows the hierarchy of evidence-based medicine, first referring to the latest clinical guidelines, then supplementing with key academic studies and cross-checking information across multiple sources.
Health encyclopedia content also undergoes multiple rounds of review and verification, with doctors from leading hospitals involved in editorial checks, according to the company.
Bringing clinical reasoning into AI
The approach also reflects how doctors themselves evaluate medical information.
Zhang Yong, chief physician of the urology department at Beijing Tiantan Hospital, said he frequently sees patients bringing AI-generated suggestions to consultations.
In his view, the strength of Qwen’s system is that it incorporates the logic of evidence-based medicine into the product: reviewing guidelines first, examining research evidence and then making a comprehensive judgment.
“By clearly listing the sources, AI answers become like food ingredient labels — users can see and verify them themselves. That is the kind of AI health assistant people can trust,” Zhang said.
A test for AI healthcare’s future
The development comes as technology companies around the world compete to apply large language models to medical education, health management and clinical assistance.
The key question is no longer only whether AI can generate medical answers, but whether those answers can be trusted, explained and independently verified.
Qwen’s “health traceability” feature represents one possible path toward solving AI healthcare’s trust deficit.
Why it matters for global audiences
For international observers, Qwen’s latest move offers a glimpse into how China’s AI healthcare sector is tackling one of the biggest challenges facing the industry worldwide: trust. China’s healthcare AI market has a unique scale advantage.
With hundreds of millions of potential users and Qwen handling tens of millions of health-related queries daily, the reliability of AI-generated medical information is not just a product issue — it is becoming a broader public health concern.
Globally, companies developing AI healthcare tools are facing a similar dilemma: how to deliver instant, personalized assistance while ensuring that recommendations remain safe, explainable and medically grounded.
By attaching sources to AI-generated advice in the same way food companies disclose ingredients, Qwen is testing a model of “accountable AI” in healthcare — where users can inspect the evidence behind an answer rather than simply trust the algorithm.
