When Healthcare Meets Capital: The Restructuring of Sovereignty Through AI, Insurance, Pharmaceutical Companies, and Platform-Based Healthcare

Over the past decade, the global healthcare industry has witnessed a power shift that has been virtually unannounced.
It wasn't voted on in any country's legislature, nor was it explicitly defined in any international treaty. It quietly unfolds in the daily transactions of capital markets, gradually accumulates in the product updates of technology platforms, slowly solidifies in the contract terms of insurance companies, and is repeatedly confirmed in the pricing decisions of pharmaceutical companies.
The result is that in many medical settings, the question of who has the decision-making power in healthcare has a completely different answer than it did twenty years ago.
This article attempts to present this shift, because understanding it is a prerequisite for understanding the most real challenges facing healthcare sovereignty today.

(Image caption) The transfer of power in the healthcare system is not declared through legislation, but is gradually completed through the daily operations of capital markets and technology platforms.

I. Insurance Capital: From Payer to Decision Maker In the traditional narrative of the medical relationship, the insurance company is a payer—it is responsible for covering the costs after the medical services are provided, but the medical decision itself is a matter of professional judgment between the physician and the patient.
This description is an outdated model today.
Through prior authorization, utilization review, and network design, large U.S. insurance companies have substantially intervened in every aspect of healthcare decision-making.
Prior authorization requires physicians to obtain approval from the insurance company before prescribing certain treatments. Approval may be denied or granted only after administrative delays. In this process, an insurance company reviewer without medical training can overturn the judgment of a physician with thirty years of clinical experience.
According to a 2023 survey by the American Medical Association, 94% of physicians surveyed said that prior authorization caused delays in patient care, and 33% of physicians said that these delays led to serious adverse medical outcomes.
Network design influences healthcare pathways in another way: insurance companies decide which hospitals and physicians are "in-network," and the cost difference between in-network and out-of-network essentially determines which healthcare resources patients can afford. This decision is based on factors including the outcome of business negotiations, strategic partnerships, and cost control targets, not just healthcare quality.
This means that in the current US healthcare landscape, insurance companies have evolved from payers into substantial participants in healthcare decision-making. This evolution has been gradual and has hardly been adequately debated in public discourse, but its consequences are real.

(Image caption) As prior authorization and utilization review become the norm, part of medical decision-making has shifted from physicians to the administrative processes of insurance companies.

II. Pharmaceutical companies: Pricing is sovereignty. The price of a drug in the United States can be three to ten times that of the same drug in Germany, and five to fifteen times that of the same drug in Canada.
This gap does not stem from differences in production costs, nor from different allocations of R&D investment—for the same drug, R&D costs are shared by the global market. It arises from differences in pricing negotiation power: the United States is one of the few developed countries in the world without a systematic drug price negotiation mechanism.
The direct result of this phenomenon is that pharmaceutical companies have a high degree of pricing autonomy in the US market, and the cost of this autonomy is borne by US patients and the insurance system.
The CAR-T therapy price of 1.2 million yuan is an extreme but representative example. This price is not based on some objective cost accounting, but on the pharmaceutical company's assessment of the market's willingness to pay, and the acceptance level of patients and insurance companies in the absence of alternative options.
In global comparisons, the UK's NICE (National Institute for Health and Care Excellence) and Germany's IQWiG have relatively well-established cost-benefit assessment mechanisms for new drugs, which can effectively constrain pharmaceutical companies' pricing during negotiations. In recent years, China has also achieved significant results in reducing the prices of some expensive drugs through national medical insurance negotiations.
However, the very existence of these mechanisms reveals the essence of the problem: the struggle for drug pricing sovereignty is one of the most financially significant battlegrounds for healthcare sovereignty. A country without pricing negotiation power has a partial deficiency in its healthcare sovereignty regarding drug accessibility.

(Image caption) Drug prices do not only reflect costs, but also pricing power. When pricing power is concentrated in the hands of a few companies, healthcare accessibility is redistributed.

III. AI: The Dual Nature of Efficiency Promises and Sovereignty Transfer
The application of AI in healthcare has been one of the most widely discussed topics in the past five years. Diagnostic assistance, image analysis, risk prediction, clinical decision support, and administrative automation—AI promises to improve efficiency, reduce human error, and enable healthcare resources to benefit a wider range of people.
These promises are supported in specific scenarios. However, the application of AI in healthcare also raises a rarely discussed issue of sovereignty: when an AI system assists or participates in medical decision-making, how is the attribution of responsibility determined?
Theoretically, there is a standard answer to this question: AI is an assistive tool, and the ultimate responsibility for decision-making still lies with the physician. However, in practice, when AI provides a diagnostic suggestion, and the physician's time pressure and trust in the system make him inclined to accept this suggestion, the definition of "AI as an assistive tool" begins to blur.
A deeper issue is the sovereignty of training data. The capabilities of an AI healthcare system depend on the data it is trained on. If that data comes from a specific population—such as a US clinical dataset primarily composed of high-income white men—then the system faces the risk of systemic bias in its applicability to other populations.
Deploying an AI system trained on data from a specific population to an environment with a completely different demographic is one of the most common, yet least publicly discussed, sovereignty issues in global AI healthcare applications today.
Whoever uses the data to train the AI, in a sense, wields influence over all users of that AI. This is not an abstract ethical issue, but a very specific issue of institutional design.

(Image caption) AI has improved medical efficiency, but it has also introduced new control logic: when decision-making relies on algorithms, the impact depends on the data source and model design.

IV. Platform-based Healthcare: When Tech Giants Enter the Healthcare Industry
Apple, Google, and Amazon—these three companies have all made significant inroads into the healthcare field in different ways over the past decade.
Apple Health integrates health data into personal devices, allowing users to store and share their medical records. This design is positive in terms of patient empowerment, but it also means that users' health data enters Apple's data ecosystem.
Google Health has advanced several medical AI projects, including AI screening for diabetic retinopathy and natural language processing of medical records. Google also owns DeepMind, whose AlphaFold protein structure prediction has become an important tool in life science research.
Amazon entered the pharmaceutical delivery market through the acquisition of PillPack, entered primary care through Amazon Clinic, and became a cloud infrastructure provider for a large number of healthcare institutions through AWS.
The common characteristic of these entries is that tech giants bring efficiency, capital, and technological capabilities, but also a very specific transfer of sovereignty. When healthcare institutions store data on AWS, when patients' health records are in Apple Health, and when the core algorithms of medical AI belong to Google, the actual control of this medical data and medical decision-making tools has been partially transferred from healthcare institutions and patients to technology platforms.
This transfer did not occur through coercion, but rather naturally through convenience and market advantages. However, its institutional consequences are real: medical data, as a strategic resource, is increasingly accumulating in the hands of a few global technology platforms.

V. AiTmed's Position: An Alternative Architecture, or Another Form of Centralization?
In this industry landscape, what are Chen Jiarui and AiTmed trying to build?
From a design perspective, AiTmed aims to establish a patient-led healthcare data architecture, using blockchain-based notarization and patient authorization mechanisms to give patients back actual control over their medical records. This direction contrasts with the centralized path taken by tech giants, and is closer to a decentralized, sovereign architecture.
But an important question needs to be asked honestly: In actual operation, can AiTmed's architecture truly achieve the patient-led approach it claims?
Any platform, regardless of its design intent, faces a gravitational pull once it reaches a certain scale: the tension between centralized business logic and decentralized sovereignty goals. Platforms need data to improve service quality, improved service quality attracts more users, and more users generate more data—the ultimate form of this cycle is often some degree of data centralization.
This is not a denial of AiTmed, but rather a structural challenge that all healthcare technology platforms must honestly confront. True patient data sovereignty requires more than just technical architecture design; it also necessitates corresponding commitments and constraints in business models, governance structures, and legal frameworks.
Dr. Chen's awareness of the problem is clear: the problem in healthcare is a systemic problem, not just a technical one. But the distance between awareness of the problem and its implementation is always more difficult to bridge than any technical solution.

(Image caption) When medical data is stored on a platform, control also shifts. Behind the convenience lies a re-centralization of data and power.

VI. Where sovereignty lies, capital flows. From GFM’s financial perspective, understanding the sovereign restructuring of healthcare capital is not just an academic issue; it directly impacts investment logic, compliance risks, and long-term asset valuation.
When a healthcare technology company controls healthcare data in a region, it effectively controls an informational asset about that population, the long-term value of which far exceeds current revenue. This explains why tech giants are willing to enter the healthcare market operating at a loss; they are not investing in current business, but in the long-term accumulation of data assets.
When a country lacks the capacity to effectively protect its domestic medical data, that data can become a strategic asset for foreign platforms. The loss of this asset may not be visible in short-term financial statements, but its long-term institutional costs can be profound.
The significance of healthcare sovereignty in the capital market is evolving from a "soft" governance concept into a risk factor with hard financial implications.
Understanding this evolution is an analytical dimension that no institution or individual making long-term investments in the healthcare field should overlook today.
In the next article, we will use all the analytical tools we have built up throughout the series to answer the final question: Can the underlying rules of healthcare be rewritten, and what is needed for such a rewriting?

Next article: "What's truly scarce is the ability to judge when to use and when to refuse."

The data cited in this article includes the American Medical Association (AMA) 2023 pre-authorization survey report and publicly available information from the healthcare industry in various countries, and does not constitute any investment advice.