Set as preferred source
Insights from State of Digital Health: Global | 2025 recap.
TL;DR
Digital health is consolidating into a capital-intensive, AI-driven infrastructure layer. With 71% of global funding anchored in the US and nearly half concentrated in mega-rounds, competitive advantage is shifting toward data control, operational maturity, and scalable execution.
Structural overview
The digital health sector has crossed a structural threshold. What was previously a fragmented, innovation-driven landscape is consolidating into an infrastructure layer embedded within healthcare systems. Growth persists, but its composition has fundamentally changed.
Funding reached $22.3B in 2025, up 19% year over year, while deal volume declined by 9%. This divergence reflects a move away from broad experimentation toward selective scaling. Capital is no longer underwriting optionality—it is underwriting control.
At the center of this shift is AI. Not as an incremental capability, but as the foundation of value creation. Models, datasets, and developer ecosystems are becoming the primary sources of differentiation. Application-layer innovation, without control of these assets, is increasingly exposed to commoditization.
The structural shift: from innovation cycles to control layers
Digital health is reorganizing around three control layers: capital, data, and distribution.
The rise of mega-rounds—accounting for 44% of total funding ($9.8B)—marks a decisive shift. Scale is no longer achieved gradually; it is front-loaded through large capital deployments. This compresses the competitive landscape and prioritizes companies that can deploy capital efficiently.
At the same time, early-stage funding has declined to 59% of total deal share, its lowest level on record. This reflects a repricing of risk. Early innovation is no longer the primary driver of value. Instead, value accrues to platforms capable of integrating data, deploying AI at scale, and controlling downstream distribution.
As fewer companies receive larger rounds, they accumulate structural advantages—data access, model performance, and distribution reach—that attract further capital. The system reinforces itself.
The market begins to resemble infrastructure rather than a traditional startup ecosystem.
Geographic asymmetry: capital as a proxy for control
The concentration of 71% of global funding in the US defines how digital health evolves globally.
Capital at this scale shapes standards, regulatory pathways, and ecosystem design. Companies operating within this environment gain disproportionate access to partnerships, distribution, and follow-on funding.
Outside the US, constraints are structural. Median deal sizes in Europe and Asia remain significantly lower, limiting the ability to scale late-stage platforms. Innovation continues to emerge globally, but scaling mechanisms are concentrated.
Over time, this creates dependency. Non-US ecosystems risk becoming sources of upstream innovation, while control over infrastructure and distribution remains centralized.
Within the US, regional clusters reinforce this advantage. Silicon Valley and Boston show higher concentrations of mid-stage companies, reflecting environments where innovation more consistently converts into scalable businesses.
AI as infrastructure: the new healthcare stack
AI has moved from application layer to infrastructure layer.
The emergence of 14 new unicorns in 2025—all AI-native—signals that value creation is now anchored in model performance and data pipelines. Clinical documentation, drug discovery, and provider workflows are becoming components of a broader AI-driven stack.
Workforce data supports this shift. The fastest-growing segments include healthcare developer platforms (+52% year-over-year headcount growth), AI model development (+34%), and robotics in care delivery (+38%). These are system-building layers rather than end-user applications.
This changes how advantage is defined. Companies that do not control models, data, or integration layers are increasingly exposed to margin pressure. The boundary is no longer between competing products, but between infrastructure owners and participants.
Data ownership: the primary asset class
The growing share of AI companies in M&A activity—nearly a quarter of all deals—signals a shift in what is being acquired.
Applications are no longer the primary target. Data is.
Proprietary datasets support model performance, regulatory defensibility, and long-term differentiation. As a result, acquisition strategies are shifting toward securing exclusive data access rather than expanding product portfolios.
This redefines competitive positioning. Brand, distribution, and product design remain relevant, but they are secondary to data ownership and model quality.
It also introduces a constraint. High-quality healthcare data is limited. As leading players consolidate access, barriers to entry increase.
Market structure: consolidation as a default outcome
M&A activity rose 33% in 2025, reaching 210 deals and reversing a multi-year decline. This reflects a structural shift rather than a cyclical rebound.
As building capabilities becomes more capital-intensive, acquisition becomes the most efficient path to growth. Companies are acquiring AI capabilities, clinical datasets, and distribution channels.
Large-scale transactions reinforce this pattern. Data platforms are no longer supporting tools—they are central assets.
At the same time, IPO activity remains limited. Exit pathways are increasingly defined by acquisition. Growth is being assembled, not just generated.
Operating leverage and emerging fragility
AI-driven models are producing high levels of efficiency. Some companies now reach valuation-per-employee levels exceeding $170M, reflecting strong operating leverage.
This efficiency introduces fragility.
When value is concentrated in small teams and highly optimized systems, performance becomes more sensitive to disruption. Changes in regulation, model accuracy, or adoption can have disproportionate effects.
The same dynamics that enable rapid scaling also increase exposure to volatility.
Risk landscape: structural, not cyclical
Several risks emerge directly from these structural shifts.
Innovation pipeline compression limits long-term adaptability. With fewer early-stage entrants, the system relies on a narrower set of innovation pathways.
Geographic concentration introduces strategic exposure. Regions without access to large-scale capital risk losing influence over infrastructure and data governance.
Data consolidation raises both competitive and regulatory concerns, as access becomes more restricted.
Exit concentration around M&A reduces flexibility, making growth dependent on acquisition timing and strategic fit.
High operating leverage increases sensitivity to shocks in a regulated and complex sector.
These risks are interconnected. They reflect a system that is becoming more efficient, but also more centralized and less adaptable.
Strategic implications
Digital health is transitioning from a fragmented innovation market into a concentrated infrastructure layer.
Capital allocation is no longer expanding the ecosystem—it is defining its boundaries. AI is restructuring healthcare systems, not just improving them. Data ownership is emerging as the primary determinant of long-term value, while geographic capital concentration shapes who controls the system.
At the same time, reduced early-stage activity introduces a constraint on future innovation diversity. Growth is increasingly tied to the ability to acquire, integrate, and scale.
The competitive landscape is narrowing. Advantage is accruing to those who control infrastructure—capital, data, and distribution—while others operate within those systems.
