AI in Oncology: Balancing Innovation With Physician Expertise

At the 2026 NCCN Annual Conference, Dr. Samyukta Mullangi, a board-certified medical oncologist with Tennessee Oncology serving patients in Nashville and Dickson, sat down for an interview with CURE to discuss the evolving role of artificial intelligence, or AI, in oncology.

During the interview, Mullangi described how AI tools may help synthesize complex clinical data, although reinforcing that physician expertise remains essential for nuanced, patient-centered cancer treatment decisions in practice.

Transcript

How might artificial intelligence support treatment decision-making in oncology, and what role should physicians continue to play when using these tools?

Yeah, so oncology is obviously very complex. It is oftentimes a chronic disease that spans months to years. It involves the incorporation of so many different pieces of data, so it's multimodal. You, as a physician, are looking at notes, genomics, formularies, patient experiences, and symptoms, a bunch of really orthogonal pieces of information to craft your decision.

I think in oncology in particular, an AI clinical decision support tool should also have access to all of those bits of information that you, the physician, are going to take into account. But frequently, it's not going to have that. There's always going to be additional pieces of info that the physician has access to in terms of experience, or what the patient is telling them, or whatever it may be.

I really do think that AI has the ability to unlock a lot of nuance in everyday decision-making, but I think the physician's judgment is going to remain paramount. I think AI is going to be a super helpful tool in oncology. But I think regulating it, because it is such a high-stakes clinical setting, is important to make sure that it's remaining a safe, assistive, and value-additive tool.