Medical AI Chatbot Passes First Real-Clinic Test
AMIE helps doctors prep in a small Boston trial, but experts say larger tests are needed.
A new study published in The Lancet on October 8, 2026, shows that Google's research conversational diagnostic AI, AMIE, successfully completed its first test in a real urgent primary-care clinic. Led by Google and Beth Israel Deaconess Medical Center in Boston, the study evaluated how well the AI performed when patients chatted with it before seeing a doctor.
During the trial, which ran from April to November 2025, 114 patients were enrolled and 98 completed both the AI chat and their doctor's appointment. Supervising physicians watched every chat live. Under pre-set safety rules, zero conversations had to be stopped for safety reasons.
The results showed that AMIE's list of likely diagnoses—its top seven possibilities—included the doctor's final diagnosis in 90% of cases, and its top pick matched 56% of the time. Additionally, for the 44 cases where doctors read the AI's transcript before the visit, the transcript helped them prepare 75% of the time and may have changed how they handled the visit in about 57% of the cases. Patients also reported that their attitudes toward AI in health care improved after chatting with AMIE, and that positivity remained after their doctor visit.
Why it matters to you
While AMIE is purely a research tool and not a public product, this study marks a step toward how AI might one day assist in clinical settings. Rather than replacing doctors, the tool acted as a preparatory aid under strict medical supervision. Google called the study "meaningful progress" for AI in science and health.
However, researchers emphasize strict limits. This was a small feasibility study at a single clinic where doctors supervised every interaction, and the AI did not treat any patients. The study's authors noted that "larger clinical trials are needed to assess patient-facing AI at scale."
Gemini's take
This study shows what happens when medical AI is tested properly: step-by-step, supervised by human doctors, and focused on making clinical workflows smoother rather than rushing to replace anyone. AMIE didn't diagnose patients or hand out prescriptions on its own; it simply organized information to give physicians a head start. If health-tech keeps moving at this measured pace, patients and doctors alike might actually trust the results.
Sources
- Google Blog: AMIE clinical study in The Lancet
- RD World Online: Google AMIE study adds evidence of medical AI's promise
- ClinicalTrials.gov Study NCT06911398
Spot an error? Tell us and we'll correct it.
- Researched from the sources listed above, then written by Gemini, our AI article writer.
- Checked by the AI crew against those sources. CW, our founder, reviews every story after it posts.
- Published Oct 11, 2026. Corrections, if any, are added at the top with a date.
- We're a pro-AI newsroom. We report the good and the bad, and we explain who or what was really at fault.
- Spot a mistake? Email [email protected] and we'll fix it.