Insights ¦ the coming evolution of healthcare AI toward a modular architecture_final

Key Takeaways

1. The proliferation of AI solutions in healthcare is creating fragmentation, prompting a shift towards a modular, enterprise-wide AI architecture.
2. Leading healthcare organisations should focus on integrating point solutions into a connected architecture that enables coordination across domains.
3. The emergence of clinical-data foundries—curated, high-quality data repositories—will unlock value by enabling advanced AI-driven insights and innovations.
4. Successful digital transformation relies on establishing strong data governance frameworks prioritising privacy, confidentiality, and risk management.
5. Private and public sector collaboration on data sharing and governance can foster the development of proprietary AI models rooted in local clinical assets.
6. The future healthcare AI ecosystem will be characterised by interoperability, orchestrated via protocols like the Model Context Protocol (MCP), and powered by agentic AI mesh.
7. Investment focus will shift from stand-alone point solutions to scalable platform architectures, with emphasis on data integration and clinical-data foundries.
8. Organisations that sequence their AI strategy from quick ROI point solutions to enterprise-wide, end-to-end domains will secure a competitive advantage.
9. Hyperscalers and major tech companies are expanding their footprint in healthcare through partnerships, data-sharing initiatives, and developing scalable AI infrastructure.
10. The competitive landscape is likely to consolidate, with EHR providers and point solution vendors competing, with solutions with proven traction retaining an edge.
11. Data assets and high-quality clinical information are emerging as key strategic enablers, translating into potential new profit pools for health systems.
12. As the healthcare landscape evolves, organisations that build capabilities in governance, integration, and data activation will be best positioned for long-term success.

Key Statistics

• Over 1,000 AI-enabled medical devices authorised by the FDA between 2015 and 2024, mostly in medical imaging.
• Approximately 2,400 healthcare-related AI companies launched within the same period, with around 1,750 backed by venture capital.
• Venture capital investment in healthcare AI increased by nearly 20% between 2023 and 2024.
• Largest funding round in 2025 was $210 million for OpenEvidence, with subsequent funding of $200 million.
• Nearly 80% of FDA-authorised AI medical devices are in medical imaging sectors such as radiology and ultrasound.
• US health expenditure reached $4.9 trillion in 2023, representing nearly 17.6% of GDP.
• Less than 40% of health systems using EHRs contribute data to de-identified database models for AI development.

Key Discussion Points

• The rapid expansion of AI point solutions addresses specific workflow challenges but risk creating operational fragmentation.
• Long-term success depends on integrating these point solutions into a modular, interconnected AI architecture to enable enterprise-wide coordination.
• Organisational strategies should prioritise end-to-end redesign of functions to support AI-native workflows rather than isolated pilot projects.
• Leadership must focus on establishing robust data governance, including data rights, sharing policies, and model accountability, to foster trust and compliance.
• The integration of agentic AI and interoperability protocols will be central to orchestrating cross-domain workflows and data flows.
• Hyperscalers are positioning themselves as key enablers by developing healthcare-specific AI platforms, cloud services, and standards for data interoperability.
• The development of clinical-data foundries will create a new revenue stream for health systems, harnessing de-identified, longitudinal clinical data.
• Future healthcare AI ecosystems will be characterised by open architectures and standardised protocols, reducing reliance on legacy data lakes.
• Investment has shifted towards workflow platforms and ecosystem-building ventures, indicating a strategic pivot from point solutions.
• Consolidation among vendors, especially between EHR providers and point solution vendors, will shape the competitive landscape, with technology differentiation becoming key.
• Building an integrated, secure, and governable data foundation is critical for scaling AI innovations and realising long-term value.
• Organisations that actively develop governance, interoperability, and data activation capabilities will be best positioned in the evolving AI-driven healthcare environment.

Document Description

This article explores the future trajectory of healthcare AI towards a modular, connected architecture. It highlights how the proliferation of point solutions has created fragmentation, and emphasises the need for integrated data infrastructure, clinical-data foundries, and interoperability protocols. The article discusses strategic implications for healthcare organisations, technology providers, and investors, including the importance of strong governance frameworks and scalable AI platforms. It underscores the role of hyperscalers and data assets in shaping a more unified, enterprise-wide AI environment that will enable innovation, operational efficiencies, and improved clinical outcomes in the healthcare sector.