Patients and Doctors Push Back on AI Replacing Physicians by 2030

A provocative article in JAMA arguing that artificial intelligence could replace physicians as early as 2030 has sparked intense debate—but the conversation may be missing the most important voices: patients and the clinicians who care for them.

The JAMA perspective by Neal Khosla and colleagues points to five cognitive tasks where AI already matches or exceeds physician capabilities, including data gathering, guideline retrieval, and documentation. Yet survey data suggests the public is moving in the opposite direction. According to a January 2026 survey commissioned by Ohio State University Wexner Medical Center, public openness to AI in healthcare fell from 52% in 2024 to just 42% in 2026, even as 51% of adults reported making an important health decision with AI assistance.

Why it matters

The gap between technological capability and patient acceptance could determine whether AI transforms healthcare or simply automates administrative tasks. With physician burnout at record levels and workforce shortages mounting, the industry faces pressure to embrace automation—but only if patients will accept it and only if it addresses the right problems.

Physicians see AI as support, not replacement

Clinicians interviewed about the prospect of AI replacement were uniformly skeptical, though not from fear of obsolescence. Dr. Roy Lirov, both a physician and computer science expert building AI infrastructure, believes the technology's greatest utility lies in reducing administrative burden rather than replacing clinical judgment.

"While the pace of improvement is head-spinning, AI replacing docs seems like a bit of a moonshot to me," Lirov said. He argues that measuring single tasks like creating differential diagnoses misses what matters in actual practice. "Anyone who's taken care of patients knows exactly how poorly standardized test scores translate to actually practicing medicine."

Of thirteen clinicians asked to respond, none believed AI could replace them in the context of actual medical practice. Their skepticism stems from clinical experience showing that medicine involves more than recall and pattern recognition—it requires intuition, judgment, and trust built through human relationships.

The debate largely rests on published studies that simulate single tasks rather than real-time encounters in exam rooms, critical care units, and surgical suites.

Trust and access concerns in underserved communities

Tamika Williams, a community health worker and founder of the Empower CHW Institute, raises concerns about how AI replacement would affect underserved rural communities already facing severe clinician shortages. Despite being pro-AI and using tools like Claude herself, Williams cannot envision technology replacing human healthcare relationships in the communities she serves.

"AI needs to be used in the right context and should not be solely relied on for human relationships," Williams said. She worries that pushing technology as a healthcare solution would undermine decades of work building trust in primary care relationships. "We would have to unlearn everything we have been preaching to communities for decades about seeking a primary care relationship. Why would they trust us then?"

Her perspective highlights a critical question often overlooked in technology development: if communities aren't included in design stages, how can the industry prevent widening existing access gaps?

The human factor

Elon Musk recently told The Economist that "there really won't be anything AI can't do better than humans, apart from being human." That exception may be precisely what matters most in healthcare.

The article concludes that while humans appear capable of building machines that can execute most healthcare functions, listening to those delivering and receiving care suggests a different path forward. Healthcare remains fundamentally a human endeavor built on human-to-human relationships—and the question of whether that foundation should be abandoned may belong to patients and clinicians, not technologists.

These findings were first reported by Lisa Fitzpatrick in Forbes.