HIMSS AI Leadership Strategy Summit: Managing Change, What's Next, and How to Plan

AI Leadership Strategy Summit: Change, Foresight, Action

At the AI Leadership Strategy Summit in Chicago, the focus was clear: manage change, scan the horizon, and build a practical plan. HIMSS President and CEO Hal Wolf emphasized that leaders need to turn AI from experiments into an operating model.

Why change management is the constraint

• Ownership: Appoint an executive sponsor and an AI Council to set priorities and clear blockers.
• Governance: Define data use, security, model risk, and human oversight. Write it down. Enforce it.
• Communication: Explain what AI will do, what it will not do, and how teams benefit. Repeat it often.
• Procurement and legal: Standardize vendor reviews for data handling, IP, compliance, and service levels.
• Adoption: Train for new workflows, not features. Tie incentives to measured outcomes.

What's on the horizon

• Agents and automation: Task-chaining across systems will move from demos to production use cases.
• Multimodal input: Text, image, and voice in one flow will compress cycle times across service, ops, and R&D.
• Model lifecycle: Monitoring, cost control, and evaluation will become day-to-day management, not a side project.
• Regulation and risk: Expect clearer rules on transparency, safety, and data retention. Prepare evidence, not slideware.

90-day plan to get ahead

• Days 0-30: Stand up an AI Council. Inventory use cases. Confirm data access, security controls, and compliance gates. Pick two pilots tied to one metric each.
• Days 31-60: Build pilot workflows end-to-end. Define evaluation sets and baseline metrics. Document human-in-the-loop steps.
• Days 61-90: Run controlled launches. Track impact vs. baseline. Decide scale, iterate, or stop. Publish an operating playbook.

Operating model checkpoints

• Strategy: Tie every AI initiative to a cost, revenue, or risk objective.
• Architecture: Confirm data pipelines, identity, logging, and isolation by default.
• Controls: Model cards, audit trails, incident response, and red-teaming baked into releases.
• People: Name accountable owners. Clarify skills, training, and escalation paths.

Workforce: skills that move the needle

• Product and process: AI product managers and process leads who can map work, not just features.
• Data and platforms: Data engineering, MLOps, model evaluation, and cost management.
• Risk and compliance: Security, privacy, and model risk with authority to pause releases.
• Frontline enablement: Training, playbooks, and change champions embedded in teams.

Metrics executives should track

• Value: Cycle time reduction, deflection rate, revenue uplift, cost per task.
• Quality: Accuracy, precision/recall, error rate, override rate.
• Adoption: Active users, task completion, time in workflow.
• Risk: Incidents, privacy events, model drift, vendor uptime.
• Unit economics: Cost per inference, infra spend vs. value created.

Vendor and build decisions

• Data treatment: No training on your data without explicit approval; clear retention and deletion terms.
• Security: Isolation options, key management, audit logging, and third-party attestations.
• Interoperability: APIs, connectors, and export paths to avoid lock-in.
• Observability: Usage, cost, latency, and quality dashboards you control.

Policy and risk resources

Use established frameworks to speed governance and evidence. The NIST AI Risk Management Framework is a solid starting point for controls and documentation. NIST AI RMF

Build capability while you execute

If you need structured upskilling for leaders and operators, review role-based programs and tool training. Start here: Complete AI Training - Courses by Job

Bottom line

The summit's message was blunt: AI is a leadership issue. Manage change with intent, look ahead with discipline, and commit to a plan you can measure. The organizations that do this will ship useful work, at lower risk, on repeat.