You've just received a $3,000 bill for an emergency room visit your insurance should have covered. Before calling the hospital's billing department, you paste the charges into ChatGPT and ask what to do next.

You're not alone. According to the New York Times, patients are increasingly using AI chatbots to decode medical bills, draft appeal letters, and identify potential billing errors — tasks that once required expensive patient advocates or hours of frustrating phone calls.

The appeal is obvious. Hospital billing is deliberately opaque, packed with codes and jargon designed to confuse. AI can translate that alphabet soup into plain English and suggest which charges to question. For patients without the time or expertise to navigate Byzantine insurance systems, chatbots offer a tempting shortcut.

But the results, as the Times reports, have been decidedly mixed. AI chatbots don't actually understand medical billing — they're pattern-matching machines that sometimes hallucinate plausible-sounding advice. They might confidently tell you to dispute a legitimate charge or miss obvious errors a human advocate would catch immediately.

The Information Asymmetry Problem

The deeper issue here isn't whether AI can help — it's that patients need this help at all. Medical billing in the United States is a deliberately adversarial system where providers and insurers hold all the cards. Chatbots can narrow that information gap, but they can't fix the underlying dysfunction.

When an AI suggests you challenge a charge, you're still the one on the phone with the billing department. If the chatbot got it wrong, you've just weakened your negotiating position. And unlike a human patient advocate, ChatGPT won't stick around to help you appeal the appeal.

The real winners here? Probably the AI companies, who get free training data every time you paste your medical records into their chat window. Read those terms of service carefully before uploading anything with your health information on it.