Why patients turn to chatbots for medical advice

OpenAI reported in July that ChatGPT answers medical questions for 300 million people each week, and one KFF survey found that about a third of people were asking AI for help with medical questions. Most people (68%), though, are still using search engines.

When KFF asked respondents why they were turning to AI, 65% reported needing “quick and immediate advice” as a major reason. Some also said they wanted to look up information about their condition before seeing a provider (41%) or that they felt more comfortable “looking up questions privately” (36%).

People have been using the internet to learn more about their health for decades now. What’s different with AI, according to Oliver Kharraz, CEO and founder of health tech Zocdoc, is that the information can be more personalized and more engaging, making it alluring to patients.

“Every time that one of these waves of more engaging media comes out…there’s maybe a risk that some patients overly trust that information,” he said. “Then the norms start setting in, where [patients] use it as a tool, and it’s frequently a good initiator of action.”

But for now, providers may find themselves frequently contending with inaccurate or incomplete advice their patients get from AI chatbots.

In a January study published in Nature Medicine, for instance, researchers used 60 clinician-authored scenarios on medical issues, ranging from sore throats to suicidal ideation, and then prompted ChatGPT to address them. Among the findings: The AI bot gave correct medical advice less than half the time for both nonurgent and emergency vignettes.

The American Medical Association (AMA) recently conducted a survey that found 70% of polled physicians said AI for general-purpose health information has a positive or neutral impact when used “appropriately.” The AMA announced in May that, in the absence of federal safeguards, it’s providing patients with information like infographics to teach them how to safely use AI.

“As technology evolves, it’s vital that patients use AI to complement—not replace—the advice of their doctors,” AMA CEO and EVP John Whyte said in a statement. “Safe and informed use of AI helps patients make smarter health decisions, but medical expertise remains irreplaceable.”

But healthcare is very expensive, the uninsured population is growing amid Medicaid cuts, and patients need more affordable options. KFF reported in May 2025 that 36% of 2,539 patients surveyed said they had skipped or postponed healthcare they needed in the last 12 months due to cost. Meanwhile, AI is just a couple clicks or taps away—and it’s free.

Kharraz noted that AI can help organize information, but it cannot replace physical and emotional aspects of human interaction.

“The way that AI and humans make decisions is very, very different,” Kharraz said.

“ChatGPT is not a doctor and should never be used as a substitute for medical care, diagnosis, or treatment,” OpenAI said in a statement shared with Healthcare Brew. “Treating chatbots as the whole story behind people’s medical decisions or outcomes oversimplifies a much bigger challenge, and risks getting in the way of people accessing powerful new tools that can aid them in their health journey.”