From telehealth to managed care: Ping An Good Doctor’s CEO on AI, insurance, and China’s ageing population

He Mingke, CEO of Ping An Good Doctor, outlines in an exclusive Q&A with HealthTechAsia how the Shanghai-based digital health platform is evolving from a large-scale telehealth service into an AI-driven, insurance-integrated managed care ecosystem.

With more than 400 million registered users, the company is now repositioning healthcare delivery around proactive health management, continuous patient engagement, and deep integration between medical services and financial protection, underpinned by large-scale deployment of medical AI, a nationwide clinical network, and a rapidly expanding model of “AI + human doctor” care.

Ping An Good Doctor has 400 million registered users. How are patients using the service differently compared to five years ago?

Five years ago, in 2021, Ping An Good Doctor had over 400 million registered users. In the same year, we advanced to Strategy 2.0, focusing on enterprise payers and developing a China-specific managed care model, thereby establishing a differentiated competitive advantage. As a result, our user services have undergone significant structural changes:

First, from “passive consultations” to “proactive health management.” Previously, users mostly came online to consult doctors when they were ill. Now, with the integration of AI and human doctors available 24/7 and proactive follow-ups, our services have shifted to the forefront, covering sub-health management and chronic disease intervention.

For healthy and sub-healthy individuals, we provide services such as health record management and medical report interpretation.

For people with chronic conditions, our health managers, nutritionists, physicians, and psychological counselors collaborate to deliver personalised, proactive health management around medication, lifestyle, medical arrangements, and more. In 2025, the number of chronic disease management service uses reached 16.9 million.

Second, from “basic online consultation” to an “online, in-hospital, in-home and in-company” health care services, we have built a comprehensive network covering “online access, hospital visits, at-home care, and enterprise services.” Users now expect more than just text-based consultations; they require holistic management covering pre-consultation triage, in-consultation companionship, and post-consultation rehabilitation.

Third, from “standalone services” to “health-insurance integration.” Users are increasingly aware that healthcare services and insurance payout are natural partners. In 2025, the add-on insurance rate among our health and senior care service users increased by 4 percentage points.

For customers using our healthcare benefits, new life insurance policies per person rose to 1.5 times the previous level; for home-based senior care service users, this figure rose to 5.2 times; and for premium senior care service users, it rose to 23.4 times.

Ping An Good Doctor holds health data on hundreds of millions of people. How do you build trust with patients around how that data is used?

As a pioneer in the Internet healthcare sector, Ping An Good Doctor has always regarded user data security as the lifeline of the company’s development. Health data involves individuals’ core privacy and is highly sensitive; the legality and security of handling such personal information are directly linked to public trust and the reputation of the industry.

Against the backdrop of deep integration between digital technology and healthcare, we have actively implemented our “CARE” sustainability strategy, built upon four cornerstones: Comprehensive Sustainability, Advanced Technology Empowerment, Robust Information Security Protection, and Extensive Social Responsibility. These pillars drive our efforts to promote high-quality, sustainable development across the industry.

In the area of robust information security protection, we have obtained a number of authoritative certifications, including ISO 27001 Information Security Management System, ISO 27701 Privacy Information Management System, and ISO 27799 Health Information Security Management System. These certifications cover all of our business areas and ensure the effective protection of data security and user privacy.

Ping An Renowned Doctor Digital Avatar connects patients to real doctors when needed. Will that always be necessary, or does AI eventually handle everything?

AI serves as an “efficiency booster,” while doctors remain the “ultimate gatekeepers.”

The value of digital the Renowned Doctor Digital Avatar lies in “breaking through the limitations of time and space,” making the expertise of top-tier hospital specialists available 24/7 and accessible to all. These digital twins can handle standardised scenarios such as report interpretation, rehabilitation guidance, and medication reminders. However, in areas such as the diagnosis and treatment of complex diseases, emotional support, and ethical decision-making, real doctors are irreplaceable.

Currently, the accuracy rate of our AI-assisted diagnosis reaches 95.1%, and the accuracy rate for complex disease MDT (multidisciplinary team) treatment plans is nearly 90%.

Yet, the final decision-making authority for diagnosis always rests with the doctor. In the future, AI will take on more standardised tasks, allowing doctors to focus on more nuanced and complex medical care that requires a human touch. This is human-machine collaboration, not human-machine replacement.

If a Ping An Renowned Doctor Digital Avatar gives a patient the wrong advice, who is accountable?

Our renowned doctor digital avatars are developed based on two foundational pillars: our “self-built network of top-tier specialists” and the “Ping An Medical Master ®” medical large language model.

By training on the actual images, professional expertise, and clinical experience of real medical experts, we have launched an AI assistant service for renowned doctors. These AI assistants provide responses within their respective specialty areas.

On the model side, distinct from general large models, the “Ping An Medical Master ®” multimodal medical large language model is built on four major medical databases (personal/corporate information database, disease directory database, pharmaceutical/medical device database, hospital/doctor database). It employs the DAGCD anti-hallucination algorithm, effectively equipping the AI system with an “intelligent radar” to accurately detect and correct output deviations.

On the validation side, continuous labeling and feedback from doctors enable ongoing iteration of the model, resulting in higher alignment with the standards required of quality physicians and more accurate completion of various tasks in medical scenarios.

On the application side, a differentiated advantage is achieved through the combination of “AI + real doctors”: AI empowers consultation services, while key steps such as diagnostics and prescription issuance are always ultimately provided and reviewed by real physicians.

AI cut your cost per consultation by 45% last year. How do your healthcare and insurance businesses reinforce each other, and what role does your senior care offering play in that model?

First, according to the Group’s annual report, in 2025 the add-on insurance rate among customers using our healthcare and eldercare services increased by 4 percentage points.

For healthcare benefits customers, the average number of new life insurance policies per customer rose to 1.5 times the previous level; for home-based eldercare benefits customers, this figure rose to 5.2 times; and for high-quality eldercare benefits customers, it rose to 23.4 times. These figures fully verify the commercial value of medical-insurance collaboration.

Second, AI empowerment and medical–insurance collaboration are two ends of the “flywheel effect”:

AI side: In the fourth quarter of 2025, with the help of AI, the per-consultation cost decreased by about 45% year-on-year. The annual number of AI doctor users approached 12 million, with AI’s contribution to gross profit reaching nearly 80 million RMB, accounting for 4.5% of total gross profit. This means we can serve a broader population at a lower cost, while providing proactive, long-term, continuous, and personalised professional services.

Collaboration side: The combined effect of cost reduction and higher-quality services further reinforces the synergistic value. For example, the “Anyouhu” (安有护) health service program provides critical illness insurance customers with end-to-end management from suspected diagnosis to rehabilitation follow-up, granting payers differentiated competitiveness and offering users with “Worry-free, Time-saving, and Cost-effective” experience, resulting in a win-win-win for all parties.

As this flywheel turns, marginal costs decrease while service quality improves; this is the core logic of managed care.

A patient in a remote county can now access a breast cancer specialist within three hours through Ping An Renowned Doctor Digital Avatar. China has over 300 million people aged 60 and above. How is Ping An reaching the people who need care most, and how far is there still to go?

Since its inception, Ping An Good Doctor has adhered to its mission of “building a professional bridge between doctors and patients and safeguarding the health of all people.” Leveraging advantages in technological empowerment, network coverage, and service system resources, the company provides the most cost-effective healthcare and eldercare services.

Technological Empowerment:
Through AI + human doctors, Renowned Doctor Digital Avatar, and multidisciplinary team (MDT) consultation platforms for complex diseases, patients in remote areas can receive expert diagnostic support for conditions such as breast cancer within three hours.

Currently, with AI empowerment, the accuracy rate of AI-assisted consultations has reached approximately 98%. AI doctors are capable of accurately diagnosing more than 11,300 types of diseases, with an overall AI-assisted diagnosis and treatment accuracy rate of 95.1%, and the accuracy rate for MDT solutions for complex diseases approaching 90%. AI-assisted doctors can handle up to 4 million consultations per day. In 2025, the number of annual users of AI doctors neared 12 million.

Network Coverage:
After more than a decade of development, Ping An Good Doctor has established a broad, high-quality, and cost-effective service network system.

By the end of 2025, we had built a team of around 50,000 internal and external doctors, signed contracts with over 3,500 specialist doctors, entered into partnerships with more than 5,100 hospitals and over 240,000 pharmacies, and collaborated with more than 4,400 health check-up service providers.

Ping An describes a managed care model with Chinese characteristics. What makes it different, and can it work in other countries?

After years of exploration, Ping An Good Doctor has developed a China-specific managed care model, enabling deep integration between “insurance payment” and “medical services.” Through proactive intervention, continuous management, and resource coordination, we help clients fall ill less often, postpone the onset of illness, and avoid serious diseases.

First, the payer landscape is diverse. In the U.S., HMOs are overwhelmingly dominated by commercial insurance, while China’s system is based on social insurance supplemented by commercial insurance. Our model connects with social insurance and, leveraging Ping An Group’s core insurance business, achieves health-insurance synergy.

We jointly participate in serving insurance policy holders through health risk management, service process management, and medical quality management, balancing the cost and effectiveness of medical services with the cost and services of insurance policies.

Second, technology-driven. Relying on the Ping An Medical Master ®（平安医博通）foundation model and our five major industry-specific models, we achieve AI-assisted triage, chronic disease management, and MDT (multidisciplinary team) consultations, resulting in improvements in quality and efficiency.

Third, integration with home-based senior care. In China, 90% of the elderly choose to age at home. We deeply integrate medical services with senior care, launching a “Ping An Concierge” service system that combines AI concierge, lifestyle concierge, and healthcare concierge features—creating an innovative advantage in the integration of healthcare and eldercare.

Ping An Good Doctor accelerates AI healthcare services and expands nationwide network in H1 2025

Which markets outside China are you in now, and where are you going next?

Can this model be expanded to other countries? From a technology perspective, our medical foundation models and medical AI applications have the potential to be replicated across regions.

However, the business model must be localised-payment systems, regulatory frameworks, and the allocation of medical resources differ greatly from country to country. At present, we remain focused on the Chinese market.

Health AI regulation looks completely different depending on which country you’re in. How does Ping An stay ahead of that – rather than just reacting to it?

In fact, since our inception, we have been continuously building, exploring, and enhancing medical AI technologies and applications. With the advancement, breakthroughs, and implementation of relevant policies, we have also been constantly iterating.

Healthcare is a serious matter, and we have always believed that medical AI must possess the core capabilities and values of “professionalism, closed-loop management, and collaboration.” An AI doctor not only needs to have highly professional medical skills, but also must provide end-to-end solutions covering diagnosis, treatment, and rehabilitation, and collaborate continuously and evolve alongside human doctors.

The vitality of an “AI doctor” lies in its continuous evolution and the establishment of a comprehensive iterative system: Each year, we complete tens of thousands of professional annotations and optimise over 10,000 cases; 50 leading experts conduct 210,000 detailed evaluations across 19 dimensions; and the model is continuously refined through ongoing user feedback and doctor annotations.

In terms of application scenarios, this is also our strong suit. Guided by enabling business value, we have developed nearly a hundred robots for four key scenarios: clinical, insurance claims, corporate health, and home care. In 2025, we will further deepen our presence in rigorous medical fields and build differentiated advantages through the collaboration of “AI + human doctors.”

On the compliance front, our AI system achieves 100% full-quality inspection in areas such as online consultation, electronic medical records, and prescription quality control, effectively safeguarding the quality of medical services.

What is Ping An Good Doctor building now that we won’t see for another few years?

There is no doubt that, with the development of the managed care model, we hope to advance the practical application of AI in serious medical scenarios through the closed-loop integration of “finance and healthcare.” Our goal is to use AI to create more revenue and enhance the quality of our services.