U.S. healthcare services disruption: Bernstein sees AI and affordability pressures reshaping U.S. healthcare services, led by care automation and value-based care

Overview

This is Bernstein's annual review of disruptive private healthcare-services companies and the structural forces affecting the U.S. healthcare system. The report argues that high medical-cost inflation, fiscal pressure, access constraints and technology advances are accelerating adoption of AI, value-based care, new pharmacy models and specialized care-delivery platforms.

Core views

Bernstein expands its annual ranking from 25 to 30 private healthcare disruptors because affordability, access, outcomes and patient-experience challenges have widened the opportunity set. The report notes that healthcare cost inflation historically ran at roughly twice CPI, with a normal pre-COVID cost trend of 6%, while employer medical-cost trends are now nearing double digits. Government programs and managed-care organizations also face high medical costs alongside low rate increases. Bernstein argues that these pressures leave payers and providers seeking structural cost reduction rather than merely passing costs through. AI is Bernstein's most disruptive long-term theme because it could change healthcare supply economics rather than only manage demand. The report separates the opportunity into care extension and automation, consumer AI agents, provider efficiencies, payer automation and efficiencies, and precision healthcare. Provider and payer efficiency applications—including ambient documentation, coding, claims processing, staffing and call-center automation—are expected to arrive sooner. Care automation and consumer agents are considered the larger eventual opportunities: Bernstein estimates trillion-dollar-plus addressable markets and profit pools above $100 billion for each, compared with $200-$500 billion addressable markets and profit pools in the tens of billions for provider and payer efficiency applications. For care extension, Bernstein assumes approximately $2.7 trillion of U.S. physician and hospital spending in 2025 and an additional 10-20% of care delivered without additional clinician compensation. This implies $250-$500+ billion of annual incremental provider revenue and more than $100 billion of annual value creation, assuming clinician compensation represents 30% of costs. The report estimates clinician expense at more than $750 billion and identifies further upside if AI addresses other clinical expenses. Adoption is expected to start in narrow functions such as radiology, then broaden as technology and regulation develop. The envisioned progression runs from human clinician aids to digital tools extending clinicians, limited-function autonomous tools, clinician-supervised "junior robot doctors," and ultimately limited forms of fully autonomous care delivery. Consumer AI agents could influence more than $5 trillion of total healthcare spending by helping patients with triage, provider selection, scheduling, cost comparison, insurance optimization, pharmacy decisions and handoffs to human care. Bernstein estimates that capturing 1-2 percentage points of savings from improved outcomes, satisfaction and affordability could create a $50-$100+ billion profit pool. The report sees agents as potentially becoming a major competitive interface for managed-care organizations and providers, although their adoption depends on accurate information, trusted clinical boundaries and effective integration into health-system applications. Provider efficiency is expected to be an earlier practical AI use case. Bernstein estimates more than $500 billion of aggregate provider administrative, operational and billing-related costs, with potential savings of roughly 15-20% through automation of documentation, revenue-cycle management, billing, staffing and workflow. Hospitals and other care providers are characterized as the biggest AI beneficiaries because lower compensation ratios could expand margins and, over time, care extension could support volume growth. The report expects larger providers to invest earlier and potentially consolidate their advantages. For payers, Bernstein estimates total operating expenses or SG&A of roughly $200-$300 billion, with about half in scope for AI after excluding areas such as taxes, fees, IT and networks. It estimates around 20% savings in managed-care operating expenses, led by care management, claims processing, call centers and other member-facing functions. However, the report expects much of the benefit ultimately to flow through to customers, leaving managed care broadly neutral at the sector level. The key stock-level differentiation would be share gains for scaled, AI-capable organizations that can lower operating costs and build effective consumer-agent capabilities. PBMs are expected to benefit from claims automation and smarter pricing, formulary and rebate optimization, though these gains may offset gross-margin pressure. Value-based care is the other central structural disruption. Bernstein argues that VBC can align financial incentives around lower total cost of care by shifting providers from fee-for-service to risk-bearing models. Established VBC companies are said to generate total savings of 15-25% by steering acute episodes to lower-cost settings, preventing costly events through chronic-condition management, closing care gaps and changing patient behavior. Bernstein sees risk adjustment as having the most immediate effect in the first 6-12 months; steerage tends to require 1-3 years, while prevention-related benefits may take 12-24 months to emerge. The report estimates full-risk VBC at about 15% of the market today and forecasts roughly 25% penetration over the next decade, with a multi-decade growth path analogous to the historical rise of managed care. Medicare Advantage is the most penetrated segment: VBC-related spend rose from 10% in 2017 to 34% in 2023, versus 12% in traditional Medicare and approximately 9% in Medicaid and 8% in commercial insurance. Bernstein expects some plateauing in 2025-26 as VBC operators become more disciplined in contracting, but sees most MA lives eventually moving into VBC by 2030 and later expansion into individual and employer markets. It estimates the VBC opportunity could reduce healthcare costs by as much as $1 trillion. Bernstein distinguishes three platform VBC models. Build models create new risk-bearing clinics but face an early loss-making J-curve as fixed costs precede patient density and savings realization. Own-and-change models acquire existing practices and convert them to VBC, avoiding the same patient-acquisition curve but requiring capital and complex clinician transition. MSO or enablement models help independent practices take risk in exchange for shared savings; they are more capital efficient but rely on acquiring and retaining partners. The report expects hybrid models to combine owned and partnered physician networks. Pharmacy disruption is another major theme. Bernstein expects specialty pharmacy to become a larger share of healthcare spending and sees policy and market forces encouraging separation of specialty pharmacies from PBMs, greater PBM fee transparency and consumer-oriented virtual prescribing-and-dispensing models. The report contrasts 2% growth for retail pharmacy with 35% growth for home delivery and specialty pharmacy and 25% growth for clinician-administered drugs. It views transparent PBMs as a response to efforts to shift retained rebates and retail spread toward explicit administrative fees. The 2026 Disruptor 30 is ranked by expected long-term disruptive impact rather than valuation. Bernstein describes the selection as primarily qualitative, using strategic themes and a qualitative matrix covering strategic drivers, differentiated advantage, market impact, traction, standalone disruption, category leadership, execution prospects, risks and management/backing. The top five are Abridge for AI clinical documentation, PANTHERx Rare for rare-disease specialty pharmacy, Aledade for VBC enablement, Thyme Care for oncology-focused specialty VBC, and Viz.ai for AI-enabled imaging and disease detection. The report also notes a pickup in IPO and M&A activity among disruptive companies, including Hinge Health's $437 million May 2025 IPO, Caris Life Sciences' $494 million June 2025 IPO, and Elevance's $2.7 billion acquisition of CareBridge. Bernstein identifies regulation and societal acceptance as the two core inhibitors to AI adoption. Autonomous care faces uncertainty over scope of practice, accountability, malpractice liability, consent and workforce redesign. Consumer agents face privacy, trust, accuracy and medical-advice boundary issues. Provider and payer deployments can be limited by fragmented data pipelines, legacy IT, workflow integration, audit requirements and staff resistance. Precision medicine requires strong clinical validation and clinician acceptance.

Analysis framework

Bernstein begins with system-wide cost, access and reimbursement pressures, then identifies the healthcare business models most likely to address them. It uses an addressable-spend and savings framework to size AI opportunities, compares the maturity and economic logic of five AI categories, and uses the historical HMO transition as a reference point for value-based-care adoption. The Disruptor 30 itself is a primarily qualitative ranking of private companies based on long-term disruptive impact rather than current valuation.

Methodology notes

• Industry AnalysisSupply-demand framework

  Healthcare cost and capacity pressures assessed through supply expansion, demand management and unit-cost reduction

  The report argues that AI care extension is distinctive because it can expand clinician capacity and reduce unit costs, rather than only steer demand or shift care to lower-cost settings.

• Other

  Qualitative disruptive-impact ranking matrix

  Bernstein ranks private disruptors using strategic alignment, differentiation, market impact, traction, leadership, execution prospects, risks and management backing, rather than current valuations.

• Industry AnalysisUpstream-Midstream-Downstream Transmission

  Value-based-care savings transmission through providers, payers, care settings and patients

  The report explains how risk-bearing providers can lower total cost through risk assessment, lower-cost-site steering, prevention and behavior modification, with savings shared across the healthcare system.

Asset mapping & comparison

Structured mapping from thesis to named assets (strengths, weaknesses, peers, risks).

• HCA Healthcare (HCA)

  Hospital provider expected to benefit from AI-driven provider efficiency and care extension.

  Strengths

  Bernstein expects provider efficiency to lower compensation ratios and support margin expansion.

  Comparison

  Hospitals and other care providers are described as larger AI beneficiaries than managed-care organizations.

  Risks

  Reimbursement pressure, slower AI adoption and regulatory constraints on clinical automation.

• UnitedHealth Group (UNH)

  Managed-care organization with potential exposure to payer automation, consumer agents and Optum provider enablement.

  Strengths

  Bernstein notes that Optum Insights may become a leader in enabling providers.

  Weaknesses

  Sector-level AI benefits may be passed through to customers, leaving managed-care margins broadly stable.

  Comparison

  Scaled and AI-leading MCOs may take share from smaller competitors with operating-expense disadvantages.

  Risks

  Execution in consumer agents, regulatory requirements and the possibility that efficiencies are competed away.

• CVS Health (CVS)

  Covered healthcare company exposed to pharmacy disruption and value-based-care themes through CVS and Oak Street/VBC initiatives.

  Strengths

  The report cites CVS's Oak Street/VBC activities as an example of public-company initiatives illuminated by private-market disruption.

  Weaknesses

  PBM and pharmacy economics face disruptive pressure from transparency and policy change.

  Comparison

  Specialty pharmacy and consumer-oriented pharmacy models are expected to grow faster than traditional retail pharmacy.

  Risks

  PBM regulation, margin pressure and changing pharmacy-channel economics.

• Elevance Health (ELV)

  Managed-care organization referenced through AI consumer-agent capabilities and VBC platform activity.

  Strengths

  The report cites Elevance's Sydney platform as a consumer-agent example and Mosaic Health as a hybrid VBC platform formed with Elevance Health and CD&R.

  Weaknesses

  Managed-care sector benefits are expected to be broadly neutral as efficiency gains flow to consumers.

  Comparison

  AI capability and scale may differentiate share winners from lagging MCOs.

  Risks

  Technology adoption, data integration, regulatory scrutiny and medical-cost pressure.

• agilon health (AGL)

  Publicly covered company exposed to the report's value-based-care theme.

  Weaknesses

  Bernstein notes that public VBC companies have generally been hurt by Medicare Advantage-related margin pressure.

  Comparison

  VBC adoption is expected to accelerate despite current public-market skepticism.

  Risks

  Medicare Advantage margin pressure and disciplined contracting requirements.

Key data

• Historical healthcare cost trend6%Normal pre-COVID medical cost trend; Bernstein says healthcare cost inflation historically ran at roughly twice CPI.
• AI care-extension addressable spend$2.7TApproximate U.S. physician and hospital care spending in 2025 used in Bernstein's scoping exercise.
• Potential annual care-extension revenue$250-$500+BBased on an assumed 10-20% increase in care delivered at the same compensation cost.
• Consumer-agent profit pool$50-$100+BBased on consumer agents sharing 1-2 percentage points of savings from nearly $5T of healthcare spending.
• Provider-efficiency savings15-20%Estimated savings from more effective administrative, operational, billing and insurance functions.
• Full-risk VBC penetration~15%Current market penetration cited by Bernstein.
• VBC market penetration outlook~25%Bernstein's estimate for the next decade.
• Medicare Advantage VBC penetration34% in 2023Up from 10% in 2017; compared with 12% in traditional Medicare, 9% in Medicaid and 8% in commercial insurance.
• VBC savings potential$1TBernstein's estimate of the possible healthcare-cost reduction from value-based care.

Impact & implications

Bernstein expects AI to improve provider productivity and margins first, while care automation and consumer agents become the larger long-term opportunities. It sees managed-care organizations as competing for share rather than retaining all AI efficiency gains, and expects value-based care, specialty pharmacy and transparent PBM models to gain relevance as cost pressures persist.

Risks

• Regulatory frameworks around clinical accountability, licensure, liability, consent and scope of practice could slow AI care automation.
• Privacy requirements, AI governance rules and fragmented provider or payer data pipelines may lengthen deployment timelines.
• Clinician workforce resistance and patient concerns over AI reliability, bias, empathy and accountability could limit adoption.
• Consumer AI agents face risks around accuracy, trust, clinical boundaries and liability when recommendations affect medical decisions.
• Value-based-care operators face contracting discipline, execution complexity and Medicare Advantage-related margin pressure.

What to watch

• The pace of AI adoption in ambient documentation, revenue-cycle management, staffing, claims processing and care-management workflows.
• Regulatory developments governing autonomous clinical functions, medical liability, patient privacy and AI oversight.
• Whether scaled MCOs convert AI efficiency and consumer-agent capabilities into share gains.
• VBC penetration in Medicare Advantage and its expansion into individual, employer and Medicaid populations.
• Policy changes affecting PBM transparency, retained rebates, retail spread and specialty-pharmacy competition.
• IPO, M&A and private-company consolidation activity among digital-health, AI and VBC disruptors.