AI × Healthcare: The Next Trillion-Dollar Industrial Revolution

Executive Summary

Over the past decade, artificial intelligence has achieved its most significant breakthroughs in internet platforms, software, and information processing. The next phase — the one most likely to transform humanity's long-term productive capacity — may unfold in an industry that is simultaneously larger, more complex, more inefficient, and more directly consequential to every human life: healthcare.

Healthcare is home to the world's most complex data, its highest-value knowledge systems, its longest development cycles, and its highest costs of failure. Against this backdrop, global demographics are deteriorating rapidly: populations are aging, the burden of chronic disease and cancer is escalating, and the healthcare workforce shortage is becoming acute. The World Health Organization projects a global shortfall of approximately 11 million healthcare workers by 2030. Simultaneously, the global population aged 60 and above is projected to grow from approximately 1 billion in 2020 to 1.4 billion by 2030, and to 2.1 billion by 2050.

The question for the healthcare industry is therefore not whether demand exists. The real question is whether the traditional healthcare system possesses sufficient human capital, time, and operational efficiency to meet that demand. AI may be emerging as a critical technology to resolve this contradiction — and this time, AI is not simply helping doctors write reports. It is progressively penetrating the entire value chain: data → diagnosis → drug discovery → clinical trials → treatment → gene editing → surgery → outcomes monitoring.

I. Structural Demand Surge: Why Healthcare's Long-Term Opportunity Is Compelling

Healthcare's greatest long-term opportunity does not originate primarily from AI. It originates from a more fundamental reality: human longevity is increasing, and disease is becoming progressively more complex.

1. Global Population Aging Is Reshaping the Structure of Medical Demand

WHO data shows that the global population aged 60 and above stood at approximately 1 billion in 2020 and is projected to reach 1.4 billion by 2030 and 2.1 billion by 2050. In other words, from 2020 to 2050, the global elderly population will more than double. Aging does not simply generate demand for more hospitals; it generates sustained, escalating demand for long-term management of the following chronic conditions:

• Cancer
• Cardiovascular disease
• Diabetes
• Neurodegenerative diseases
• Renal disease
• Chronic respiratory disease
• Orthopaedic conditions

2. Non-Communicable Diseases Are the World's Largest Healthcare Burden

WHO data shows that non-communicable diseases caused at least 43 million deaths in 2021, representing approximately 75% of all non-pandemic-related global deaths. Of this total, cardiovascular diseases accounted for at least 19 million deaths, cancers approximately 10 million, chronic respiratory diseases approximately 4 million, and diabetes and related kidney diseases over 2 million.

This means the largest future healthcare market is not the treatment of a single disease — it is the continuous understanding of the human body, the early detection of disease, the precise selection of treatment pathways, and the long-term monitoring of therapeutic outcomes. This is precisely where AI has one of its greatest potential advantages: processing vast datasets, identifying complex patterns, and generating predictions.

3. Healthcare Supply Cannot Keep Pace With Demand: Workforce Gap Is the Core Bottleneck

One of the most acute bottlenecks in the global healthcare system is human capital. Physicians, nurses, radiologists, pathologists, drug researchers, and clinical trial personnel all require years of specialist training. WHO currently projects a global healthcare workforce shortage of approximately 11 million by 2030. AI will not produce more doctors in the near term, but it can materially improve the productivity of existing physicians and healthcare systems. The genuine value of AI in medicine is therefore more likely to manifest as enabling one physician to handle the workload that previously required several, rather than replacing physicians outright.

II. AI Enters Healthcare: The Opportunity Is a Complete Value Chain, Not a Single Product

The traditional healthcare value chain has been highly fragmented across its nodes: patient → testing → physician diagnosis → drug discovery → clinical trials → treatment → surgery → long-term monitoring. AI is progressively integrating these segments, forming a closed loop from data to treatment. The most important competitive dynamic in healthcare AI will not necessarily be which company possesses the most powerful model, but which company can command the complete data–experiment–treatment feedback loop.

III. FDA Data Confirms: Healthcare AI Has Entered Real Clinical Systems

One of the most important methods of assessing whether a technology delivers genuine value is to observe whether regulators have begun to approve it. The US FDA's AI medical device database provides a critical observation window. As of the end of 2025, the FDA had authorized 1,357 AI/machine-learning medical devices across 693 companies and 17 clinical specialty areas. The cumulative total surpassed 1,000 authorizations in 2024, with 258 new authorizations granted in the first three quarters of 2025 alone. AI is actively entering real clinical settings in radiology, cardiovascular medicine, neurology, ophthalmology, surgery, pathology, ultrasound, and oncological diagnosis. This is no longer a laboratory concept.

IV. The First Frontier: Medical Imaging and Diagnosis

Medical imaging is fundamentally a large-scale pattern recognition problem. CT, MRI, X-Ray, ultrasound, and pathology slides are all instances of the same process: data → image → pattern → judgment. This is one of deep learning's most natural domains of competence.

A study published in JAMA Network Open found that as of June 2024, the FDA had authorized 950 AI/machine-learning medical devices, of which 723 (76%) belonged to radiology. A separate study published in npj Digital Medicine analysed 1,016 FDA authorizations as of 2024 and found that images remained the predominant AI input type, accounting for 84.4% of cases.

This establishes an important point: the first phase of AI's entry into healthcare is likely not the 'AI physician' — it is AI becoming the physician's second pair of eyes. Stanford's 2025 AI Index summarised a broad body of medical AI research, with some studies finding that advanced AI models can match or exceed physician performance on complex clinical diagnostic tasks. Crucially, a growing body of evidence suggests that AI working in conjunction with physicians typically outperforms either AI alone or physicians alone. This means the more realistic future competitive dynamic is not AI replacing physicians, but physicians who are most proficient with AI gradually displacing those who are not.

The more realistic competitive dynamic is not AI replacing physicians, but physicians who are most proficient with AI gradually displacing those who are not.

V. Beyond Imaging: AI Begins to Process Biology Itself

Medical imaging represents only the first phase. The truly transformative industrial opportunity lies in AI beginning to process biology itself. The human body encompasses DNA, RNA, proteins, cells, metabolism, the immune system, microbiome, pharmaceuticals, environment, and clinical manifestations — interacting with extraordinary complexity. Scientists have historically struggled to process so many variables simultaneously. AI offers this capability for the first time.

AlphaFold: AI Opens the Data Architecture of Life Science

In 2024, Google DeepMind's AlphaFold project was awarded the Nobel Prize in Chemistry. This was not a routine AI headline — it signified that AI had begun to solve biological problems that traditional computational methods had long struggled to address efficiently. AlphaFold has predicted over 200 million protein structures; its database is used by over 3 million researchers across more than 190 countries.

AlphaFold 3 has further extended its scope to model the interactions between proteins, DNA, RNA, small molecules, ions, and other biological molecules. This points to a fundamental shift in drug discovery: in the past, scientists searched for molecules; in the future, AI may first understand disease mechanisms and then design the molecules accordingly.

VI. AI Drug Discovery: Transforming Healthcare's Most Expensive Bottleneck

Traditional drug development presents a profound economic challenge. A new drug typically requires more than 10–15 years to develop, with R&D costs often exceeding US$1–2 billion, and approximately 90% of candidates entering clinical-stage development ultimately fail. A primary cause of failure is not the absence of drug candidates, but the inability to identify those that are genuinely efficacious.

AI seeks to alter this process fundamentally. The traditional pathway proceeds through: disease → target identification → high-throughput molecular screening → laboratory experiments → optimisation → animal studies → clinical trials. The AI pathway flows through: large-scale biological data → AI identification of disease mechanisms → prediction of potential targets → generative molecular design → AI-predicted properties → automated experimentation → clinical trials. The greatest potential value is not making scientists 20% faster — it may be eliminating the large volume of candidates that should never have entered the laboratory in the first place.

The First Real-World AI Drug Discovery Case: Insilico Medicine

Insilico Medicine (03696.HK) is a closely watched case study. Its AI platform was involved in the development of the TNIK inhibitor ISM001-055 for idiopathic pulmonary fibrosis. The journey from discovery to drug candidate took approximately 18 months, compared to the roughly 4–5 years a traditional development pathway would require. The programme has completed Phase IIa and reported positive efficacy results. The significance of this outcome is material: the industry's earlier sceptical question — 'Can AI actually produce real drugs?' — is progressively giving way to a more advanced question: 'Can AI consistently produce successful drugs?' These are fundamentally different questions.

AI Drug Discovery: Core Company Profiles

Recursion Pharmaceuticals (RXRX): Its core proposition centres on AI + automated laboratory + large-scale cellular data, aspiring to construct what might be described as a machine for automated drug discovery. Key advantage: substantial proprietary experimental data. Key risk: clinical success rates remain inherently uncertain.

Insilico Medicine (03696.HK): Its core proposition is generative AI + drug design + laboratory automation. It is one of very few AI-native pharmaceutical companies listed on public markets. The critical validation milestone is whether AI-designed drugs can be brought to commercial products.

Schrödinger (SDGR): This company pursues a distinct approach: physics + molecular simulation + AI, engaging more directly with the underlying physical mechanisms of drug molecules. Its technology has advanced multiple drug candidates into late-stage clinical development, illustrating how computational chemistry is transitioning from a tool into core drug development infrastructure.

VII. mRNA: One of the Most Important Bridges Between AI and Therapeutics

If AI can understand disease, the logical next step is designing treatments. mRNA occupies a critical position in this progression. Conventional drugs manufacture a fixed molecule; mRNA is closer in nature to transmitting a manufacturing instruction to the body's own cells. This means that when a virus generates new mutations, the theoretical workflow becomes: AI analyses the virus → designs a new mRNA sequence → rapidly manufactures a vaccine → enters clinical testing. This is why COVID-19 gave the world its first widespread understanding that mRNA is a platform technology, not merely a vaccine type.

Moderna and BioNTech: From Vaccines to Oncology

The genuinely compelling long-term research case for Moderna (MRNA) and BioNTech (BNTX) is no longer COVID-19 — it is cancer. The prospective pathway is: tumour sequencing → AI identification of tumour-specific antigens → personalised mRNA design → immune system training → targeted attack on cancer cells. This suggests that AI + genomic sequencing + mRNA + immunotherapy could converge into an entirely new paradigm for oncology treatment.

VIII. Gene Editing: From Managing Symptoms to Modifying the Source of Disease

If mRNA transmits instructions to the human body, gene editing directly modifies DNA. CRISPR Therapeutics (CRSP) has demonstrated that CRISPR-based gene editing can enter the commercial stage. Its Casgevy therapy was among the first globally approved CRISPR gene editing treatments, indicated for sickle cell disease and transfusion-dependent beta-thalassaemia. Gene editing has moved from scientific experiment to real clinical medicine.

Beam Therapeutics: Next-Generation Base Editing

Beam Therapeutics (BEAM) represents a distinct direction: Base Editing. Conventional CRISPR functions like scissors. Base Editing is closer to correcting a single letter within the DNA sequence — offering a higher degree of precision. The clinical data from BEAM-302 in 2026 warrants close attention. Company reports indicate that at the 60mg dose, mean steady-state total AAT reached 16.1 μM, all patients sustained levels consistently above the 11 μM protective threshold, mutant Z-AAT declined by 84%, and normal M-AAT represented 94% of total AAT. The company plans to initiate a global pivotal study in the second half of 2026. If it is subsequently demonstrated that a single in vivo Base Editing intervention can produce durable correction of a genetic disease, the gene editing market may enter an entirely new phase of development.

IX. Oncology: The Most Important Battleground for Healthcare AI

Cancer commands particular significance because it is not a single disease — it is a collection of hundreds of distinct diseases. Different patients present with different genomes, different mutations, different tumour microenvironments, and different immune responses. Precision medicine combined with AI is therefore a natural fit for oncology. The future cancer value chain may take shape as follows: blood/tissue → genomic sequencing → AI mutation identification → cancer risk prediction → treatment selection → personalised therapeutics → treatment → ctDNA testing → recurrence assessment. This architecture will generate a cohort of critically important companies.

Natera: The Molecular GPS for Post-Treatment Cancer Surveillance

Natera's (NTRA) core product, Signatera, addresses MRD — minimal residual disease. Following cancer surgery, the absence of visibly detectable cancer cells does not confirm that cancer cells have been eradicated. Through ctDNA testing, Signatera can detect tumour DNA present at extremely low concentrations in blood. This may fundamentally alter cancer management: the historical standard of periodic CT imaging could give way to continuous molecular-level surveillance.

Guardant Health: Migrating Cancer Detection From Tissue to Blood

Guardant Health (GH) is centred on liquid biopsy. Conventional methods require tissue extraction; the future requires only a blood draw, followed by analysis of DNA, RNA, and epigenetic information. In 2026, the FDA approved the updated Guardant360 Liquid CDx, with its detection scope significantly expanded, further reinforcing the role of liquid biopsy in precision oncology. Cancer diagnosis is progressively transitioning from 'detecting a tumour' to 'understanding a tumour.'

Tempus AI: The Core Platform for Healthcare AI Data

Tempus AI (NASDAQ: TEM) operates with a logic that is meaningfully distinct from pure-play diagnostics companies. It is attempting to build a comprehensive healthcare data platform integrating clinical data + genomic data + pathological data + AI. Future value may be generated through the closed loop of: data → AI → clinical decision-making → drug discovery → new data. This is one of the most important business models in healthcare AI: data is not a one-time asset but an infrastructure asset that can continuously train AI and continuously generate new value.

X. Robotic Surgery: AI Eventually Enters Physical Intervention

If data is AI's brain and pharmaceuticals are AI's therapeutic instrument, then robots are AI's hands. Intuitive Surgical (NASDAQ: ISRG), through its Da Vinci surgical robotic system, represents the most commercially mature case study in robotic surgery. Its competitive strength is not derived from selling robots; it lies in the complete commercial closed loop formed by robots + physician training + surgical data + consumables + hospital ecosystem. AI may progressively extend further into surgical path planning, tissue identification, risk prediction, surgical motion analysis, automated assistance, and physician training. The trajectory of robotic surgery may shift from 'physician operating the robot' to 'physician supervising AI-assisted robot.'

XI. The Genuine Long-Term Opportunity: A Five-Layer Value Chain Framework

Many investors encountering AI healthcare instinctively search for the 'AI physician.' This may not represent the largest opportunity. The genuine long-term value is more likely distributed across the complete value chain: Layer 1 — Data (TEM/NTRA/GH); Layer 2 — AI Models (healthcare foundation models, biological models); Layer 3 — Drug Discovery (RXRX/03696/SDGR); Layer 4 — Therapeutic Technologies (BNTX/MRNA/CRSP/BEAM); Layer 5 — Robotics (ISRG); Layer 6 — AI-driven complete healthcare closed loop.

Why Healthcare AI May Become the Next Major AI-Era Industry

Healthcare simultaneously satisfies four structural conditions for deep AI penetration:

• ① Massive economic scale: global healthcare expenditure is among the largest industries in the world
• ② Exceptionally high information density: DNA, RNA, proteins, imaging, medical records, pharmaceuticals, and clinical trials are all data
• ③ Exceptionally high labour costs: physicians, researchers, nurses, and clinical trial personnel all require substantial human capital investment
• ④ Exceptionally high failure costs: a single failed drug programme can consume hundreds of millions of dollars and years of effort

Consequently, if AI can improve efficiency by even 10%, the resulting economic value could be substantial.

XII. Key Risks: Healthcare AI Is Not an Obstacle-Free Highway

Risk 1 — Data Quality. Healthcare data is highly fragmented: different hospitals use different formats, different countries operate under different privacy regulations, and different patients present different disease manifestations. If AI model training data contains systematic biases, the resulting outputs may be erroneous.

Risk 2 — Insufficient Clinical Validation. AI accuracy in laboratory settings does not translate automatically to effectiveness in real clinical environments. Despite the rapid increase in FDA-authorized AI medical devices, researchers have noted that real-world validation remains inadequate. A JAMA Network Open study found that of the 950 FDA-authorized AI/ML devices as of 2024, 723 were in radiology; among radiology devices with complete submission documentation, only 29% included clinical testing and only 5% underwent prospective testing. 'FDA authorization' does not equate to 'AI demonstrated to be effective across all real-world clinical settings.'

Risk 3 — AI Cannot Eliminate Clinical Trial Failure. AI can identify targets, design molecules, screen patients, and optimise trials, but the human body remains the ultimate test. AI cannot guarantee that a drug will be efficacious once administered to humans. A 2026 study by the Tufts Center for the Study of Drug Development showed that AI has begun helping to improve the efficiency of cancer clinical trials, with related analysis estimating that AI can shorten late-stage trials by approximately 10 weeks and reduce direct operating costs by up to approximately US$5.6 million. However, the study simultaneously noted that AI cannot resolve all clinical trial challenges — patient recruitment, informed consent, and drug supply issues persist.

Risk 4 — Regulatory Uncertainty. The regulatory framework governing healthcare AI is itself evolving rapidly. The FDA has issued guidance principles relating to AI drug development and continued to refine its regulatory framework for AI in drug and biologic development in 2026. Simultaneously, the FDA is exploring how to evaluate generative AI medical devices, as these systems may produce variable, non-deterministic outputs — fundamentally unlike traditional medical devices.

The enterprises that develop genuine and durable competitive moats in this space will therefore likely not be those with the most technologically impressive AI systems, but those that combine AI capability + clinical validation + FDA regulatory experience + proprietary healthcare data + commercialisation competence into a single coherent platform.

XIII. Competitive Evolution: From Model Competition to Data and Closed-Loop Competition

The competitive dynamics of healthcare AI will evolve through four phases: Phase 1 — whose model is more intelligent; Phase 2 — whose dataset is larger; Phase 3 — who possesses real-world clinical data; Phase 4 — who can convert AI outputs into drugs, therapies, and revenue. Ultimately, the competitive moat of healthcare AI companies will not consist of a single algorithm but of the complete closed loop formed by data, models, experiments, clinical validation, regulatory expertise, and commercialisation capabilities.

XIV. Five Investment Themes for the Next Decade

Complete Company Map

XV. Conclusion: The Self-Reinforcing Loop of the AI Healthcare Value Chain

The question of whether AI will enter healthcare has, in practice, already been largely answered. The FDA has authorized 1,357 AI medical devices; the adoption of generative AI by healthcare institutions is transitioning from pilot programmes to scaled deployment. McKinsey's 2026 survey found that approximately half of US healthcare institutions have implemented generative AI, and more than 80% have deployed at least one application to end users.

The real question has now become: to what extent will AI restructure the healthcare industry's value chain?

2026–2028, AI-Assisted Healthcare: Improvements in physician efficiency across imaging diagnosis, medical records, clinical trials, and drug screening.

2028–2032, AI Participating in Medical Decision-Making: AI begins to meaningfully participate in drug design, patient selection, personalised treatment, cancer monitoring, and surgical assistance.

2030s, AI-Driven Healthcare: AI ceases to be merely a tool and becomes the infrastructure connecting data, biology, pharmaceuticals, treatment, and surgery.

Ultimately, the healthcare industry may transition from its historical model of 'disease detection → disease treatment' toward 'disease prediction → disease prevention → precision treatment → continuous monitoring → proactive intervention.' The most profound change will not simply be a reduction in healthcare costs. More significantly, AI may for the first time enable healthcare to transition from large-scale standardised treatment to large-scale personalised medicine. This is why AI × healthcare may, over the next decade, become one of the most compelling long-term technology industries to track — ranking alongside the internet, cloud computing, and AI infrastructure as a generational structural opportunity.

AI may for the first time enable healthcare to transition from large-scale standardised treatment to large-scale personalised medicine — one of the most profound structural shifts of the next decade.