Accenture Global Health Lead on Scaling AI in Healthcare with Governance and Intent

Why It Matters

Without strong governance and measurable outcomes, AI deployments risk regulatory backlash, wasted spend, and eroded clinician confidence, limiting the technology’s potential to improve patient care and reduce costs.

Key Takeaways

• Governance and accountability are essential for scaling AI in health systems
• Workflow integration, not model performance, determines AI success
• Consolidating pilots into enterprise AI platforms improves measurement and ROI
• FHIR APIs and data normalization cut AI fragmentation costs
• Prove value in high‑friction workflows before scaling AI

Pulse Analysis

AI is reshaping health care, but the rush to adopt sophisticated models often outpaces the industry’s ability to govern them. In an interview with Healthtech Analytics, Accenture’s Andy Truscott emphasized that AI must be treated as a regulated capability, not a vanity project. He noted that patients are receptive when AI improves access, yet clinicians remain skeptical without clear accountability frameworks. This tension underscores the need for a governance layer that validates, monitors, and retires AI solutions, turning experimental tools into reliable enterprise assets.

Operational hurdles dominate the scaling journey. Truscott pointed out that most AI failures stem from poor workflow integration—models sit on the side of processes instead of inside them. Data fragmentation further erodes value, as inconsistent standards force costly data‑cleaning efforts. He also warned that organizations often purchase AI faster than they can define validation and escalation protocols, especially for evolving agentic systems. The remedy lies in consolidating disparate pilots into unified AI platforms, assigning both business and clinical owners, and leveraging interoperable standards like FHIR APIs to ensure seamless data flow.

For health systems ready to move beyond hype, Truscott recommends a disciplined, evidence‑driven rollout. Start with high‑friction, high‑impact use cases, quantify hours saved, errors reduced, and dollars recovered, then expand only after clear ROI is demonstrated. Investing in interoperability and a dedicated operating model for AI governance will lower compute costs and build clinician trust. In this measured approach, AI becomes an accelerator of efficiency rather than a risky experiment, positioning early adopters as industry leaders rather than fast‑but‑flawed adopters.