We also have X and podcasts
Why AI Healthcare Investment Follows Money, Not Medical Need
A comprehensive analysis of 3,807 AI health startups reveals stark geographic concentration, funding gaps, and critical underrepresentation of clinical expertise in founding teams.
Background
Researchers analyzed AI health startups founded 2010–2024 using PitchBook, LinkedIn, and Crunchbase data, applying manual expert coding, GPT-4 classification, and external validation to map investment patterns, geographic distribution, and founding team composition.
Key Findings
- Two-thirds of investments concentrate in clinical decision support ($10.4B), drug discovery ($18.5B), and imaging/diagnostics ($11.87B)—all higher-complexity deep-learning domains
- Mental health, public health, and rehabilitation remain significantly underfunded
- Geographic concentration: US leads with 1,609 startups and $38B; Africa and South America show minimal activity
- Founding teams predominantly technical (35%) and business-oriented; clinicians represent less than 5% of founder-only teams
- Gender representation critically low at 15.8% female founders overall, 10.7% in technical roles
- FDA-approved AI devices are 97%+ assistive/perceptual; advanced and autonomous systems remain rare (<1% each)
Why It Matters
Investment concentration overlooks clinically underserved populations. The minimal clinical expertise in founding teams suggests many AI solutions may not adequately address real-world workflows, creating an adoption gap between innovation and clinical practice.
Limitations
GPT-4 classified the full dataset after manual coding of only 400 startups. Geographic and gender metrics depend on LinkedIn and Crunchbase data, potentially underrepresenting regions with limited digital infrastructure.
Original paper: Mapping AI startup investment and innovation in healthcare using a five-tier AI systems complexity framework. — NPJ digital medicine. 10.1038/s41746-026-02595-5
