AI Is a Powerful Companion, Not a Replacement
- AI tools are increasingly becoming part of how patients learn about chronic lymphocytic leukemia (CLL), but this learning method can introduce confusion or misinformation; physicians are often finding themselves clarifying or correcting what patients read online.
- Research shows mixed performance from general‑purpose chatbots: Some answers are accurate, but many are problematic, missing context, or overly confident — making doctor‑guided tools like SurvivorNet’s My Health Questions a safer option.
- CLL treatment is highly individualized, shaped by each patient’s genetic markers, health status, and personal goals — a level of nuance AI cannot fully capture, which is why precision medicine remains “an art in the science.”
- AI can help patients feel informed and empowered, but it cannot replace the human judgment, reassurance, and personalized decision‑making that guide CLL care and ensure treatment aligns with each patient’s needs and values.
Before addressing AI at all, Dr. Yaghmour says his first priority with a new patient is to meet their emotional needs and ensure they don’t feel alone.
Read MoreHe describes this as a gradual process, adding that patients might only absorb a fraction of what’s discussed in any single visit.
Where AI Fits — And Where It Falls Short
Dr. Yaghmour admits that AI tools are now part of the picture for many patients.
“What they see on the AI technology, our information can add to, but it can sometimes confuse them, and we have to make sure we clarify,” he says.
Dr. Yaghmour’s cautious approach is well-founded by current research, though the picture is genuinely mixed.
Earlier studies, including research published in the peer reviewed medical journal Cancer Spectrum, says that of AI chatbots and cancer information found reassuring results.
One peer-reviewed study found that ChatGPT’s answers to common cancer myths and misconceptions were highly accurate compared with the National Cancer Institute’s answers, though reviewers noted the chatbot’s language was indirect, vague, and, in some cases, unclear. More recent research paints a more concerning picture.
A 2026 study in the medical journal BMJ Open shows that many mainstream chatbots still aren’t ready for medical decision‑making because researchers found that nearly half of AI‑generated answers to health questions were “problematic,” often due to misinformation, missing context, or overly confident but incorrect explanations.
A Safer Way to Use AI: Tools Built With Doctors
One example of a more reliable approach is SurvivorNet’s doctor‑supported AI tool, My Health Questions.
Unlike general‑purpose AI platforms, this tool is specifically designed for cancer patients and grounded in medical expertise. SurvivorNet’s reporting highlights how patients are using it to get clearer, more trustworthy answers to critical questions about their diagnosis, treatment options, and side effects.
Since “My Health Questions” is built with physician oversight and tailored to oncology topics, it helps reduce the risk of misinformation and gives patients a safer way to prepare for conversations with their care team.
Precision Medicine as ‘Art in the Science’
Part of why AI-generated general information can only go so far, according to Dr. Yaghmour, is that CLL treatment is highly individualized.
“We do design precision medicine. It’s personalized, case by case,” he says. “It’s like literally an art in the science because every case has its different piece that you put together for the best outcome.”
This reflects how CLL is actually treated: decisions are shaped by a patient’s specific genetic and molecular disease features, alongside their overall health and personal preferences — a level of individualized nuance that general-purpose AI tools, which respond the same way to similar questions from any user, aren’t designed to capture.
Dr. Yaghmour frames the human element of this as essential. “Patients need to have that kind of feeling that you’re not chasing numbers, you’re not there for the screen or the computer, or we’re human.”
A Message for Patients Leaning on AI
AI can help patients feel informed, organized, and empowered. It can make complex cancer information more understandable and accessible. However, it can also misinterpret, oversimplify, or inadvertently frighten patients, especially when used to analyze personal medical data.
Dr. Yaghmour’s overall message is that AI can be a helpful starting point for patients trying to understand their CLL diagnosis, but it can’t replace the individualized judgment, reassurance, and real-world context that a physician provides — including catching AI’s mistakes, tailoring treatment to each patient’s specific disease and goals.
Learn more about SurvivorNet's rigorous medical review process.
