How to perform an Oncology AI Readiness Assessment

AI adoption is accelerating faster than readiness.

As oncology organizations enter 2026 generative AI tools are widely piloted. Agentic capabilities are emerging. Yet many practices cannot clearly articulate how AI-generated inputs are governed, how AI-influenced decisions are owned, or how AI adoption advances patient support objectives tied to outcomes, cost and experience.

My last post focused on the need to add “strategy” and “governance” to the common project components of “people, process and technology.” This piece discusses what an oncology specific AI readiness assessment should test, how to translate readiness results into 5-year AI goals and how to convert those goals into a sequenced project plan across people, process, technology, strategy and governance. The focus is not on tools. It is on operational capability.

The Core Tension: AI Activity without Decision Clarity

Most oncology leaders no longer ask whether AI will be used. The more consequential question is whether their organizations are prepared for AI to influence prioritization.

Generative AI lowers the cost of producing content. Agentic AI lowers the cost of deciding what work happens next. In oncology, that distinction matters because patient support functions are constrained by time, staffing and coordination rather than insight alone.

Navigation, triage and post-discharge workflows fail when prioritization is delayed or inconsistent. Agentic AI directly addresses that constraint. It also introduces new governance requirements that many organizations have not yet defined.

Why Readiness Matters for Oncology Patient Support

Oncology patient support is the operational backbone of value-based and clinical quality performance.

Rapid symptom escalation reduces avoidable utilization. Effective post-discharge follow-up stabilizes patients. Navigation capacity determines whether patients access supportive care before crises emerge.

AI can strengthen these functions only if readiness extends beyond technical deployment. Without clear ownership, escalation rules and accountability, AI outputs become noise rather than leverage.

AI Readiness and Implementation Strategy go Hand-in-Hand

A Useful AI Readiness Framing

AI readiness in oncology should be assessed as an operating capability, not a tool inventory. A readiness assessment should be performed first because the scope of an AI strategy will be constrained by the skills and resources available to the implementation team.

Effective assessments test conditions for AI-mediated prioritization:

• Whether data reflects clinical reality

• Whether workflows specify ownership and timing

• Whether roles are redesigned rather than overlaid

• Whether outcomes align with value-based and clinical quality objectives

• Whether governance defines AI autonomy boundaries and accountability

  
  

Organizations that score poorly in any of these domains should not progress from generative to agentic AI use, regardless of vendor sophistication.

A snapshot view of an AI Readiness Assessment could look like this:

Translating Readiness into Five-Year AI Goals

A five-year AI plan should begin with defined “Triple Aim” outcomes and work backward.

The goal is not maximum automation. It is a reliable decision flow that expands capacity without eroding trust.

Early phases of agentic AI adoption should focus on assistive and advisory use cases with explicit human oversight. Orchestrated workflows should be introduced only after escalation logic, auditability and accountability are proven in practice.

Implications for Oncology Leaders

AI readiness is no longer a technical exercise. It is governance of prioritization.

Practices that approach AI incrementally without redefining decision ownership risk compounding early errors as agentic capabilities expand. Those that anchor readiness to patient support, value-based outcomes and clinician trust are better positioned to use AI as a capacity engine rather than a source of hidden risk.

If your organization is evaluating how to perform an AI readiness assessment, define five-year AI goals and translate them into a sequenced implementation plan for navigation, triage, post-discharge care and supportive services.

This is the point at which external perspective is often most valuable.

I support oncology practices with readiness assessments, governance design, workflow redesign and board-level alignment to ensure AI adoption advances patient outcomes, cost performance and clinician experience rather than undermining them.