Patient AI Use and the Decline of Physician Trust, With Robert Shpiner, MD

Physicians are increasingly on the receiving end of a healthcare system patients no longer fully trust. They see it in patients who arrive with chatbot screenshots instead of triage notes, in evaluations delayed because an AI tool or online influencer normalized symptoms that should have prompted a call, and in algorithmic reassurance standing in for clinical assessment.

The scale of the shift is documented. A 50-state survey found that the share of US adults reporting a lot of trust in physicians and hospitals fell from 71.5% in April 2020 to 40.1% in January 2024. A related analysis found that decline concentrated along educational and political lines, with a 12-point drop between 2019 and 2024 among adults without a college degree.

In a paper titled "Why Patients Ask the Chatbot First," published in Annals of Internal Medicine, Robert Shpiner, MD, of the David Geffen School of Medicine at UCLA, argues that the relationship between falling trust and rising chatbot use is not simple cause and effect. Still, the pattern is difficult to dismiss: as confidence in physicians has declined, low-friction digital tools have filled the space patients once reserved for a clinician.

In the paper, Shpiner argues that trust operates less as a feeling patients carry toward medicine and more as something built or broken by the mechanics of getting care, from appointment access to how quickly a portal message gets answered.

In the following Q&A, Shpiner discusses what is fueling the broader erosion of trust in health care, how artificial intelligence use fits into that pattern, and what internal medicine physicians can do to rebuild confidence with patients who no longer extend it by default.

Q&A: Patient AI Use and the Decline of Physician Trust, With Robert Shpiner, MD

Patient Care Online: What motivated you to write about this, and how does it connect to the broader trust problem in medicine right now?

Shpiner: I'd seen a lot of vilification of the technology, and a lot of conflict around the threat of AI to my profession. Those concerns may all be true, and I'm not saying there shouldn't be regulation — in parallel, I think there needs to be increasing oversight, because it is a wild west, and the more regulation there is, the better. But on top of that, I wanted to take the chance to ask: okay, that's true, but what are we doing as a profession? What is our contribution? What can we do to recognize and alter our behavior in a way that helps stave off, ameliorate, and regain trust? It may be a push to regain their trust, because if we're now with them on this, as opposed to against it, it may be a step toward increasing trust.

The new head of the AMA — and I love him for this, it's one of the things I've been a huge proponent of — has said the biggest threat to health care in the United States right now is the lack of trust, and that growing schism. If I were to identify a single illness, a malady, a problem — and I have the unique perspective of having been practicing for almost 40 years — that erosion, that chasm, is even more profound to me now.

Everywhere I look, the infosphere is contributing to it. Sadly, I don't know if I can say this without getting in trouble, but the government, RFK, HHS — they're all contributing to it. When they start to bring up vaccine questions and relitigate them, it causes a schism. When they push peptides through, or want to investigate whether the pneumococcal vaccine is actually effective, it sends a completely antithetical, mixed message about trust, and people get confused: "So I should take peptides, but I have to be questioning my vaccine?" That's one of the pressures.

Then there's social media, where someone says, "My friend's leg fell off two weeks after they got their pertussis vaccine," and suddenly people are into conspiracy thinking. There's a huge amount in the environment driving that schism, and it's a trust schism.

When people come into my ICU now and I start talking to them about end-of-life issues, I've had experiences where they look at me and say, "The hospital just wants to save money," or "Why are you experimenting on my mother?" Things I wouldn't have heard in the past. I don't remember that degree of inherent mistrust at the beginning.

When I walk into a room in the ICU, I can often tell by the look on families' faces that I've already lost, in terms of establishing trust, and it never used to be like that. That was always one of the skills I was able to bring: even without knowing all the answers, making the patient feel "trust me, we're going to do the best thing, everything we can." You could tell in a moment whether there was that connection. Now, there's often already a wall up before I walk in the room — "What are you doing to my mother? Why are you experimenting on her?" — across every part of the spectrum, whether it's "you're not doing enough" or "you want to pull the plug."

I think if we sit and think about that, as physicians we need to do everything we can to combat this plague of mistrust. The chatbots, the AI — they're one manifestation of it. But the bigger disease is distrust itself.

References:

1. Perlis RH, Ognyanova K, Uslu A, et al. Trust in physicians and hospitals during the COVID-19 pandemic in a 50-state survey of US adults. JAMA Netw Open. 2024;7:e2424984. doi:10.1001/jamanetworkopen.2024.24984

2. Alsan M, Cutler DM. Prescription for division: healing the growing gap in physician trust. JAMA Health Forum. 2025;6:e256765. doi:10.1001/jamahealthforum.2025.6765

3. Shpiner RB. Why patients ask the chatbot first. Ann Intern Med. doi:10.7326/ANNALS-26-01819