From Pilots to Platforms, How State Health Agencies can Build Trustworthy AI

State health agencies are under increasing pressure to adopt AI quickly. The real risk is not moving too slowly; it is building AI on foundations that were never designed for trust, scale, or accountability. Across states, promising pilots stall because data is fragmented, governance exists outside the system, and interoperability is treated as a one-off integration exercise rather than an architectural discipline. AI does not hide these gaps; it exposes them.

The agencies that succeed with AI will do so not by chasing tools, but by modernizing the data and interoperability layers that make AI viable in the first place.

FHIR® and TEFCA Signal a Structural Shift

FHIR® and TEFCA are often framed as regulatory milestones. In practice, they represent a structural change in how health data is expected to move across ecosystems.

FHIR® establishes shared semantics for structured, versioned, and meaningful data across clinical, administrative, and population health contexts. TEFCA adds the trust framework, defining how organizations exchange data at scale, under consistent rules, with accountability built in.

ASSYST’s work supporting national interoperability initiatives, including ONC’s Interoperability Standards Advisory and conformance testing for electronic prescribing and Real-Time Prescription Benefit, demonstrates how these standards operate in real environments. These platforms enable providers, payers, pharmacies, and Health IT vendors to exchange data reliably and in real time, strengthening the foundation for nationwide interoperability.

For AI, this matters. Models cannot reason across systems if meaning, provenance, and access rules change at every boundary. AI does not replace interoperability; it depends on it.

Architecture Is the Real AI Strategy

AI forces a new architecture conversation. Modernization is no longer about replacing systems one by one; it is about designing platforms where data, policy, analytics, and operations are aligned from the start.

• Next-generation health architectures share common traits
• FHIR®-native data layers instead of downstream normalization
• Embedded governance, lineage, and consent enforcement
• Reusable data products shared across programs
• Analytics and AI operate on the same governed foundation.

ASSYST’s Hephaestus platform reflects this architectural approach. Informed by interoperability and regulatory programs, Hephaestus treats interoperability, governance, and analytics as first-class capabilities. Rather than adding AI on top of brittle integrations, it enables AI to operate on trusted, standardized data that already meets policy and compliance expectations.

AI Is Changing Interoperability and Data Dissemination

AI is reshaping how health data is consumed. Leaders increasingly expect natural-language access to governed data, contextual insights delivered inside workflows, predictive and prescriptive signals rather than static reports, and transparency into how conclusions are reached.

This raises the bar for data dissemination. Governance cannot live solely in documentation or review boards; it must be enforced at runtime. Data quality, access controls, lineage, and auditability become operational requirements as AI begins to influence decisions across care delivery, public health surveillance, and program integrity.

ASSYST’s experience integrating AI into interoperability, cybersecurity, and data platforms across CMS, FDA, CDC, and HRSA demonstrates how this can be done responsibly, with AI augmenting human judgment while remaining explainable and defensible.

Why ASSYST’s Federal Experience Matters to States

Federal health programs operate at a national scale, under continuous oversight, and in accordance with evolving standards. Systems supporting Medicare, Medicaid, public health surveillance, regulatory submissions, and prescription services must function across organizations, withstand audits, and adapt as policy changes.

ASSYST’s decades of experience in these environments provide state health agencies with something increasingly valuable: proven architectural patterns under real scrutiny. From FHIR®-based interoperability and TEFCA aligned exchange to electronic prescribing, Real-Time Prescription Benefit enablement, and AI-driven data operations, this experience helps states move faster with confidence and less risk.

It also helps Health IT leaders ask the right questions early, before AI investments become costly to unwind.

Build the Rails First

AI will transform state health programs, but only if it is built on data that is interoperable, governed, and trusted by design. The winning strategy is not faster experimentation; it is intentional architecture.

Build the rails, align to standards, embed governance. Then let AI move fast, responsibly, transparently, and in service of better outcomes.

ASSYST Healthcare IT

ASSYST is a trusted Health IT Systems Integrator with over 30 years of experience advancing public health and regulatory missions. We deliver secure cloud modernization, AI-infused automation, and interoperable data platforms that enhance care delivery and operational resilience. Our solutions apply AI and analytics to prevent fraud, waste, and abuse, ensuring integrity, efficiency, and trust across the nation’s healthcare ecosystem. 

ASSYST Hephaestus is a FHIR-native Application Platform as a Service (aPaaS) evolved from the Green Accelerator Program, enabling healthcare organizations to modernize legacy data into the latest HL7® FHIR® standards. It delivers an end-to-end, microservices-based interoperability stack—FHIRSpy, FHIRBricks, FHIRProof, FHIRWire, FHIRBI, and FHIRBreak—to automate data ingestion, mapping, validation, exchange, and analytics across the healthcare ecosystem.