AI reconstructs consumer-side healthcare
```
There used to be a prevailing bias in the pharma investment circle: investors generally believed that digital healthcare was a pseudo-concept.
In their eyes, there is an “impossible triangle” in medical demand: low frequency, high barriers, and non-standardized products. Ordinary people don’t go to the doctor every day, doctors require ten years of training, and everyone’s medical records are unique.
This leads internet healthcare platforms to be trapped for years in the quagmire of buying traffic—very high customer acquisition costs, very low user retention.
However, this stereotype is being ruthlessly shattered by the data flood of the AI era. The demand for AI healthcare on the consumer side is actually astonishing; it has always existed, only lacking a sufficiently low-threshold, low-cost, and smart enough interaction container to receive it.
When the interaction cost drops to zero and the feedback quality reaches near-professional level, this silent, rigid demand for AI healthcare erupts instantly.
This explosive moment is seen in the actions of two AI supergiants: one is China’s Ant Group, the other is America’s OpenAI.
In China, Ant Group’s “Afu” has seen its monthly active users double to over 30 million in a single month, with daily queries exceeding 10 million.
Across the Pacific, OpenAI officially launched OpenAI Health on January 7, 2026.
OpenAI’s data shows that globally, more than 230 million people per week consult health questions on ChatGPT – and this was happening even before this product was launched. This demand is overflowing, it is urgent.
Ant Afu and OpenAI Health both chose to create a separate entry point—this stems from their precise grasp of consumer healthcare demand. Medical data is sensitive, it needs physical-level isolation, finance-grade security, and must make users feel safe enough to upload even their most private medical records.
Ant Afu and OpenAI Health have formally confirmed the end of the era when search engines dominated medical information, and the beginning of the era when intelligent agents are taking over personal health.
01 Ant “Afu” Demand Verification
In December 2025, Ant Group officially upgraded its AI health app to “Ant Afu.”
But subsequent data far exceeded everyone’s expectations. In just one month after the brand upgrade, Ant Afu’s monthly active users skyrocketed from 15 million to 30 million, with daily average queries exceeding 10 million.
Behind this data are two key industry insights.
First is the breakthrough in full-link ecosystem.
Ant Afu is no longer just a Q&A box; it has integrated with smart devices from ten major brands including Huawei, Apple, and OPPO, linking hardware data with “health mini-goals” and achieving end-to-end coverage from daily monitoring to online consultations and offline medical visits. It has connected with 5,000 hospitals nationwide and 300,000 real doctors, making AI not only able to “chat” but also “deliver”.
Second is the validation of huge demand in lower-tier markets.
Data shows that 55% of Afu’s users come from third-tier and below cities. In first- and second-tier cities, people may easily visit top hospitals, but in underserved markets where medical resources are scarce, people fervently desire a free “expert” to answer health questions at any time. Ant Afu directly addresses this pain point: using AI technology, it equalizes the gap in healthcare resource distribution, turning infrequent serious medical consultations into high-frequency health companionship.
02 OpenAI’s Logic for Independent Entry Points
On January 7, 2026, OpenAI officially released ChatGPT Health.
The most striking change is that OpenAI decided to set up a separate entry for Health in the sidebar.
Why do this? The core logic lies in the conflict between large AI models and privacy/security.
OpenAI knows well: users need efficiency and creativity for coding or copywriting, but for health consultations, what they need is absolute safety and privacy. If users worry that their medical records could be used to train AI, or fear that during a work demo the AI will suddenly spit out suggestions about private illnesses, they will never upload real health data to AI.
Therefore, OpenAI designed a nearly physical-level isolation architecture.
At the storage layer, conversations, files, and data in the Health space are entirely stored separately from the main interface. Health has an independent memory system—these memories will never “flow back” to the main chat. This means if you consult about a psychological condition in Health, but switch to a coding demo in the main window, AI will never leak any related information.
Even more crucial, OpenAI clearly promises: Health conversation data will not be used to train its foundational models. Only with this kind of trust can top medical institutions cooperate, and users feel secure enough to upload their genetic reports.
The independent entry is in fact a “trust firewall” OpenAI built for healthcare AI.
03 AI Healthcare Function Matrix
ChatGPT Health’s ambition goes far beyond being just a chatbot; it aims to manage the full lifecycle of user health through powerful ecosystem connections.
Currently, medical data tends to be highly fragmented, scattered across hospital EMRs, paper reports, smartwatch apps and various vertical applications. ChatGPT Health does not try to negotiate interfaces hospital by hospital; commercially, that’s not viable.
It chose a key strategic partner—b.well Connected Health.
b.well, as one of the largest real-time connected health data networks in the US, has built infrastructure based on the FHIR standard. Through this partnership, ChatGPT Health solves the problem of large models “not understanding” messy medical data.
With user authorization, users can one-click pull medical records from various hospitals; AI can interpret structured lab results, as well as deeply understand unstructured clinical notes and discharge summaries.
Beyond static medical records, ChatGPT Health also integrates with Apple Health to access dynamic physiology data, giving AI advice a time dimension.
When users complain of “palpitations,” ChatGPT can immediately pull HRV (heart rate variability) data from the past 24 hours, combine with user’s history, and determine if an urgent hospital visit is needed.
In addition, OpenAI has brought in partners like Instacart and AllTrails, bridging the final gap from “advice” to “action.”
AI can use your metabolic data to generate a diet plan and convert it directly to an Instacart shopping list; it can also, based on your fitness level, recommend suitable hiking trails via AllTrails. This closed-loop ability—from data aggregation to action—is a dimensional upgrade over traditional internet healthcare.
December’s update of Ant Afu has similar features, further bridging the gap from daily health Q&A to online consults and offline visits—proving the commonality of consumer AI healthcare needs in East and West.
04 Reshaping the Valuation Logic of AI Healthcare
From Ant Afu to OpenAI Health, the moves by China and America’s tech giants signal a fundamental shift in industry competition logic.
First is the overhaul of traffic entry points.
For two decades, users got health information mainly via search engines, monetizing via keyword ad bidding. This model is inherently conflicted, leading to mixed information quality. OpenAI Health represents a new entry form—conversational service. If users get used to getting precise, data-driven answers from AI, content platforms like WebMD and Baidu Health will quickly lose value, and traffic will irreversibly flow toward AI agents with private data moats.
Second is the revaluation of offline services.
In the AI era, algorithms themselves are cheap—you can buy compute power; but only high-quality real-world data is scarce. Offline service providers will shift from “human service” to “data asset” vendors.
Finally, pricing “trust”.
OpenAI Health’s emphasis on privacy isolation and no-training commitment is actually pricing trust. In the AI era, trust is the most valuable currency. Users dare upload their most private medical records to OpenAI because they believe in its privacy architecture. This trust will be the core premium for OpenAI Health’s commercialization (premium subscriptions, insurance partnerships) in the future.
The future Afu or OpenAI Health will be a super health butler, online 24/7, aware of all your physiology, able to mobilize real-world resources—becoming a driving force in transforming the AI healthcare ecosystem.
This article comes from WeChat Official Account “Hard AI”. For more cutting-edge AI news, go here
Risk Warning and DisclaimerThe market has risks, investment requires caution. This article does not constitute personal investment advice, nor does it consider individual users’ specific investment goals, financial situations or needs. Users should consider whether any opinions, views or conclusions contained herein are suitable to their particular situation. Investments made accordingly are at your own risk. ```