Should We Be Surprised Patients Are Turning to AI?
AI adoption has accelerated rapidly as new tools have become widely available, and people have begun experimenting with them in every corner of their lives.
For a while, I assumed individuals and organizations were adopting these tools at the speed of trust. But what has become clear is that patients are adopting them at the speed of desperation.
As access challenges deepen, affordability barriers grow, healthcare deserts expand, and mistrust in the medical profession reaches historic levels, patients are increasingly turning to AI for answers. According to a 2026 KFF survey, 32% of adults now use AI chatbots for health information or advice, despite 67% saying they trust these tools “not too much” or “not at all.” That gap between use and trust is not a sign of empowerment alone; it’s a sign of need.
And that need is not evenly distributed. Early evidence shows that AI is disproportionately used by those facing the steepest barriers to care: uninsured individuals, people of color, and patients with lower socioeconomic status—groups already navigating the heaviest burden of social determinants of health. AI can help reveal the social forces that shape health, but it also mirrors them, meaning the patients most desperate for answers are often the least well‑served by the tools they are turning to.
So, is this the new frontier of patient empowerment, autonomy, and the informed healthcare consumer? Or is it a signal that healthcare delivery must do better?
The answer is both.
Patients taking an active role in their health journey should not be viewed as a problem. Their willingness to research, question, and seek clarity is a strength. But the fact that many are turning to AI because they cannot access timely, affordable, trustworthy care is not. It is a warning that our system is not meeting people where they are.
So, what should we do?
First, healthcare must determine what aspects of AI’s emergence we should embrace, shape, and push back on.
We should embrace informed patients and do so without dismissing the active role they are taking in their care.
We should shape the clinical encounter now that patients often arrive with an AI‑generated diagnosis or care plan. We must find ways to integrate this reality while preserving the identity, judgment, and expertise of our caregivers.
And we must push back on the conditions that drive patient desperation. No one should have to rely on an algorithm because they cannot rely on their healthcare system.
At The Outcomes Institute, we are shaping these conversations and taking a systems‑level approach to the future. We cannot repeat the mistakes of the electronic health record era, where clinicians were left out of the design and forced to absorb the consequences. The pace of change may be faster outside the care walls than inside them, but the answer is not to replace human judgment, connection, and empathy with AI.
The future must be AI + Human Work™—designed deliberately, equitably, and with clinicians at the center.