From Pilot to Practice: Scaling Microsoft Copilot Safely Across Your Health System

The gap between AI experimentation and operational deployment is the defining challenge in healthcare AI today. Research published in the New England Journal of Medicine by Microsoft and The Health Management Academy found that 43% of health system executives report piloting or testing agentic AI, yet only 3% have deployed agents in live workflows. A full one-third of respondents indicate no plans to explore agentic AI within the next one to two years. 

Yet confidence in AI's long-term value is strong. 60% of respondents agree that agentic AI will meaningfully improve or disrupt the provider-patient experience, and 77% expect AI agents to improve backend productivity. The gap between conviction and execution is what Microsoft calls the "readiness divide": some health systems are building governance, security, and workforce models to scale AI safely, while others remain in proof-of-concept mode, producing diverging outcomes in productivity, workforce strain, and resilience. 

Delay carries its own costs. As Microsoft noted: "Waiting to act can mean higher costs, slower throughput, and greater strain on already-stretched teams". But scaling without governance invites regulatory penalties and reputational damage. The answer is a governed acceleration approach: assess, remediate, deploy in phases, measure, expand. 

A Phased Approach to Scaling Copilot 

Scaling Copilot across a health system is best approached in three stages: 

1. Phase 1: Controlled Pilot. Start small and safe. Roll out Copilot to a limited set of low-risk users and data. Monitor outputs closely. Use this phase to identify unexpected behaviors or policy gaps while the stakes are manageable. 
   
   
2. Phase 2: Hardening and Expansion. Address any security or compliance gaps uncovered in Phase 1: tighten access controls, adjust DLP rules, provide additional staff training. Once the environment is hardened and leadership is comfortable, extend Copilot to more users or a second use case. 
   
   
3. Phase 3: Enterprise Deployment. Integrate Copilot into everyday workflows across the organization. At this stage, AI becomes a standard operational tool. The key is operationalizing responsibly: ongoing training, clear documentation of AI's role, continuous security audits, and performance reviews. 

This phased approach aligns with Microsoft's broader framework for healthcare AI: 

• Start with workflows, not technology. Identify the highest-friction moments (documentation backlogs, member service, security triage) and design AI interventions that measurably reduce time, effort, and risk. 
• Get the foundation right, early. Prioritize secure access, identity, and data governance so Copilot has the right context without compromising privacy or compliance. 
• Make it real, and make it stick. Operationalize responsible AI (oversight, evaluation, human-in-the-loop), measure quality and safety, and invest in change management so adoption scales beyond early enthusiasts. 

Waiting to act can mean higher costs, slower throughput, and greater strain on already-stretched teams.

What Successful Scaling Looks Like 

Intermountain Health provides a strong model. After beginning a physician pilot in mid-2025, Intermountain moved to a full rollout embedded in Epic workflows. During the 13-week deployment, 894 clinicians received in-person training, 496 joined online classes, and 67 support staff attended "train the trainer" sessions. By the end of 2025, Intermountain had over 2,500 active users and was continuing to expand. 

The critical enabler was the combination of leadership commitment, structured training, specialty-specific optimization, and partnership with Microsoft and Accenture to drive lasting engagement. As Intermountain's Chief Clinical Officer JP Valin, MD noted: "As we integrate artificial intelligence and advanced data, this tool will simplify the care we deliver and improve the quality of care and experience for our patients and our communities". 

Addressing the Reskilling Imperative 

A consistent finding from the research is that scaling AI is as much a people challenge as a technology challenge. 60% of healthcare executives cite reskilling and upskilling as a top challenge as AI ecosystems expand. Leaders increasingly view agentic AI as a strategic end state that depends heavily on progress in workforce readiness, governance, and data infrastructure. Moving from promise to sustained value requires deliberate, coordinated investment across all three. 

Healthcare organizations scaling Copilot should plan for: 

• Structured training and change management: Do not assume adoption happens organically. Build programs that include specialty-specific workflows and real-time coaching. 
• Leadership alignment: Ensure CIOs, CISOs, CMIOs, and operational leaders share a common vision for AI's role. The research reinforces that success depends on how effectively organizations prepare their foundations and empower their people to work alongside AI tools. 
• Continuous improvement: Track time savings, user satisfaction, error rates, and compliance metrics. Use these data points to justify expanded investment and refine your approach. 

The GDS Readiness Roadmap 

GDS recommends a five-phase approach moving from initial assessment to full enterprise enablement: 

1. Cyber Hygiene: Implement MFA for all users; secure identities and devices for immediate risk reduction. 
2. Pay the Tech Debt: Migrate remaining legacy systems; clean up inactive accounts and data; modernize into Microsoft 365. 
3. Data Governance: Classify and protect data with Microsoft Purview sensitivity labels and DLP; tighten external sharing; set retention policies. 
4. Introduce Copilot: Enable Copilot for pilot users with guardrails; monitor usage and results. 
5. Copilot Enablement: Roll out to the broader organization with ongoing optimization, user training, and ROI tracking. 

The GDS AI Readiness Assessment serves as the starting point, providing a measurable evaluation of your current maturity across security, governance, data hygiene, workflows, and operational structure, along with a prioritized action roadmap to close gaps. 

For sustained operations, GDS's Managed Microsoft 365 services deliver continuous security monitoring through a 24x7x365 SOC, device compliance enforcement, data governance tuning, and monthly optimization reviews. As your Copilot footprint grows, your security posture grows with it. 

The Window Is Closing 

Healthcare has a rare window to define how AI shapes clinical and operational work before patterns harden and competitive advantage is locked in. Building strong governance frameworks, establishing a trusted data foundation, and developing an AI-ready workforce are no longer optional; they are prerequisites for leadership in the next era of healthcare transformation. The organizations best positioned to lead are those that know where they stand before they begin, scale deliberately, and invest in the governance and partnerships required to sustain AI at enterprise scale.  

Benefits of Managed IT Services from Global Data Systems

• Strategic Managed IT: We help you solve your technology related business problems.
• Connectivity: We get you reliable, secure connectivity anywhere in the western hemisphere in 48 hours.
• Support: When you need help simply call our 24x7x365 support number.
• Billing: Instead of managing hundreds of vendors - get one, easy to read bill from GDS.

Contact Managed Services Provider, Global Data Systems >