In a recent conversation with ACCC, DiMe Associate Program Director Ian Miller discussed what responsible AI adoption looks like in practice, particularly for organizations early in their AI journey and seeking clearer guidance on how to evaluate, prioritize, and implement these tools.
Artificial intelligence (AI) is no longer a future concept in health care. It is already embedded in scheduling tools, documentation systems, and patient communications. Yet for many cancer care professionals, the growing drumbeat around the adoption of AI raises as many practical questions as it does possibilities.
Where does AI truly belong in cancer care today? And how can oncology leaders move forward without compromising trust, wasting limited resources, or pursuing AI without a strategic plan in place?
Digital innovation, when thoughtfully applied, carries real promise. The Digital Medicine Society (DiMe), a global nonprofit and professional home for digital medicine, was created to advance the safe, effective, and equitable use of digital approaches to improve health care delivery and outcomes. DiMe’s work reflects a reality familiar to health care leaders: The success of digital tools depends less on the technology itself and more on how it is put into practice.
In a recent conversation with ACCC, DiMe Associate Program Director Ian Miller discussed what responsible AI adoption looks like in practice, particularly for organizations early in their AI journey and seeking clearer guidance on how to evaluate, prioritize, and implement these tools.
“Digital innovation really promises to redefine health care and solve some of the most pressing and persistent challenges to good health for all,” Miller said.
That promise comes with a reality many cancer programs know well: Moving too quickly without readiness can create more challenges than progress.
Despite growing interest in artificial intelligence, the challenge for many oncology programs is not whether AI is possible, but how to transition from exploration to sustainable, meaningful use. Progress often slows when tools are introduced before organizations are prepared to support them.
Some cancer programs are addressing this challenge by making an intentional commitment to start small. New tools are tested with a limited group—often, a single physician or small team—before expanding more broadly. This measured approach allows organizations to learn, adjust, and build internal confidence while protecting clinicians from unnecessary burden as AI capabilities evolve.
DiMe’s work on AI implementation recognizes that successful AI adoption moves beyond acquiring the technology alone and depends largely on organizational readiness and human-centered execution. Meaningful progress requires systems that can support new tools responsibly, particularly in oncology, where decisions are complex, care delivery is tightly coordinated, and the consequences of missteps can be significant.
When AI tools are poorly aligned with clinical practice, lack a well-defined purpose, or arrive without appropriate governance structures, they can erode clinician confidence, create workflow disruptions, and strain limited resources. In cancer care, those consequences affect care teams, patients, and the systems in place to support them.
In contrast, organizations that invest time upfront in preparation and validation are better positioned to introduce AI in ways that support clinical judgment rather than disrupt it. As Ian recalled from the Playbook, “This type of preparatory, foundational, work to de-risk, validate, and build AI literacy ultimately speeds up adoption and scale of AI solutions. Setting thresholds and guardrails early on prevents costly reconfigurations and disruptions in ‘live’ workflows, and increases trust in clinicians deploying these tools on the front line.”
Public conversations about AI often center on its potential role in diagnosis or treatment decisions, where the promise and risk can feel highest. DiMe encourages oncology programs to look elsewhere first, focusing on areas where organizations can build experience, assess value, and establish guardrails before AI intersects directly with clinical decision‑making.
As Miller emphasized, “High‑volume administrative tasks—those not clinically focused and touching patient care—are most ripe for demonstrating impact and showing that AI tools are worth your dollars.”
For cancer programs navigating staffing shortages and mounting administrative burden, task-based AI applications provide a use case and a lower risk entry point. They allow teams to become familiar with AI, establish oversight processes, and assess performance—without placing patient safety or clinical trust at risk.
AI’s long‑term potential lies not only in automation, but also in its ability to generate insight—when data can be collected, synthesized, and meaningfully interpreted in ways that support better understanding, not just faster processes.
“With the power of AI, we can collect real‑time, real‑world data, synthesize it, and understand its impact. That's so valuable to all kinds of stakeholders in the system,” Miller highlighted.
In oncology, these capabilities can help inform operational decisions, guide resource allocation, and identify opportunities to improve patient experience. DiMe leaders emphasize that insight only becomes useful when AI is intentionally deployed, continuously monitored, and grounded in clearly defined organizational goals.
The Playbook: Implementing AI in Health Care
Recognizing how often organizations struggle to move from interest to impact, DiMe developed The Playbook: Implementing AI in Healthcare. The Playbook is not a buyer’s guide or a push toward rapid adoption. It’s an interactive tool that outlines a structured decision-making process by identifying meaningful problems, assessing organizational readiness, and determining whether AI is appropriate at all. Only after establishing that foundation does it guide organizations through tool evaluation, implementation planning, and ongoing governance. This approach makes The Playbook particularly useful early in the adoption conversation, when teams may feel pressure to act but lack a shared framework.
For oncology programs, ACCC developed Assessing AI Readiness in Oncology: A Companion Checklist to be used alongside The Playbook, helping organizations translate broad AI considerations into practical planning steps across clinical, operational, and administrative teams. Additional resources, including ACCC’s tip sheet, Getting Started With Generative AI in Cancer Care Delivery and Operations, can further support internal discussions as organizations move deliberately from questions to decisions.
AI adoption in cancer care does not need to be fast to be meaningful. For many organizations, the more valuable work happens earlier—building shared understanding, using practical tools to guide evaluation, and establishing a clear approach before AI becomes part of day‑to‑day care delivery.
Learn more about ACCC’s AI in Cancer Care initiative at accc-cancer.org/ai.
