The Prepared Patient: Patients Ask AI Before They Asked Their Doctor
A physician friend told me about a recent visit that stayed with her. A patient sat down, pulled out her phone, and explained that before coming in she had typed her symptoms into an AI chatbot. She arrived with a printed list of questions, a few possible explanations, and several treatment options she wanted to talk through. A decade ago, my friend said, she might have braced herself. That day she felt something closer to hope.
The patient was doing exactly what I want every reader of The Prepared Patient to learn to do: showing up informed, curious, and ready to participate in her own care.
Artificial intelligence has arrived in the exam room whether we invited it or not. Surveys now show that a large share of patients have turned to tools like ChatGPT and Claude to interpret symptoms, decode lab results, or make sense of a diagnosis they did not fully grasp when they left the office. This is the new front door to health information, and our patients are walking through it every day.
So the question facing physicians has changed. It is no longer whether patients will use AI. It is who will guide them to use it well. I believe that responsibility belongs to us.
Why physicians are the right guides
Patients can find information almost anywhere now. What they cannot easily find is judgment and answers informed by context: the ability to weigh a piece of information against their specific history, medications, risk factors, and values. AI can generate a confident paragraph about chest pain. Only a physician who knows that this particular patient has a family history of early heart disease can tell her which sentence in that paragraph actually matters for her.
We also understand how medicine works as a system. We know when a result is urgent and when it can wait, which symptoms warrant a call at midnight or trip to the ER, and how often the textbook answer fails to fit a real human being. That contextual knowledge is exactly what AI lacks, and it is exactly what makes our guidance worth seeking.
Where AI genuinely helps
When patients use these tools thoughtfully, good things happen.
AI can help a patient prepare sharper questions before a visit, which makes our limited time together far more productive. It can translate dense medical language into plain words, so a discharge summary stops reading like a foreign document. It can help someone organize a complicated medication schedule, track symptoms over time, or rehearse a hard conversation about goals of care. For patients who feel intimidated in medical settings, and there are many, AI offers a low-pressure place to ask the question they were too embarrassed to raise out loud.
These are the moments when technology levels the playing field. A patient who once felt voiceless can walk in prepared, and preparation changes everything about how a visit unfolds.
Where patients need to exercise caution
The message every patient should hear is simple: AI is a starting point and never the final word. A few cautions matter most.
These tools can sound authoritative while being wrong. They sometimes invent facts, reference studies that do not exist, or offer guidance built on outdated information. They do not know a patient's full medical record, so their suggestions can miss a critical drug interaction or a contraindication that any clinician would catch. They are not bound by the privacy protections that govern a clinic, so personal health details typed into a public tool may not stay private. And they cannot perform an exam, order the right test, or take responsibility for what happens next.
The risk I worry about most is delay. A reassuring answer from a chatbot can talk a frightened person out of seeking care they urgently need. No algorithm should ever sit between a patient and a timely call to their physician.
And we should advise our patients NOT to upload their medical record into general LLMs. They cannot and do not ensure privacy and who knows how the companies use your information.
The physician as translator and partner
Here is where our profession can lead. We can teach patients to carry what AI tells them into the conversation instead of letting it replace one. We can show them how to ask better questions of these tools, how to notice a confident answer that deserves a second look, and how to recognize the moments that call for a human rather than a machine.
When a patient arrives with an AI-generated list, a physician sees an opening. She can correct what is wrong, affirm what is right, and use the visit to build something the technology will never provide on its own: a relationship grounded in trust.
The prepared patient is engaged, informed, and unafraid to take part in their own care. AI, used wisely and guided by the people who know medicine best, can help more patients become exactly that. The technology is here. Our patients are already using it. Let us be the ones who help them use it safely.
The Prepared Patient: Your Guide to Surviving the Health Care System (Johns Hopkins University Press, for release August 11) is about empowering patients with information, leveling the playing field between patients and the system, and helping people become active partners in their own health and health care. Follow this newsletter for more on how to prepare, ask, and advocate for yourself and the people you love.