Physicians Find ChatGPT Inaccurate for IVF Advice

A new study reveals that AI-generated responses regarding IVF often contain misinformation compared to professional medical materials.

Updated on Oct. 30, 2026 in Pregnancy

Gouache-painted illustration of a glowing geometric shape inside a glass beaker, representing the intersection of digital information and clinical advice. — New research indicates that ChatGPT frequently provides inaccurate information about IVF, often misrepresenting success rates compared to professional medical resources. AI Illustration. Upload story photo >

Live Poll

Would you trust medical information provided by an AI chatbot over professional medical society documents?

Physicians report that ChatGPT provides less accurate and safe information on in vitro fertilization than official American Society for Reproductive Medicine (ASRM) educational resources. The study found that half of the AI's answers contained unsupported claims.

Why it matters

As patients increasingly turn to AI for complex medical guidance, researchers warn that reliance on chatbots for reproductive health can lead to exposure to misinformation and inaccurate success rate estimates. Ensuring the safety and completeness of AI information is critical for those navigating fertility treatment.

Physicians rated ChatGPT’s accuracy at 3.98 out of 5 compared to a perfect 5.00 for ASRM documents. Furthermore, 4.2% of chatbot responses contained direct misinformation, while ChatGPT failed to correctly estimate live birth rates for various age brackets.

The players

American Society for Reproductive Medicine

This is a multidisciplinary organization dedicated to the advancement of reproductive medicine through education, research, and advocacy.

University of Pittsburgh

This is a public research university where the lead researcher involved in this fertility technology study is based.

The details

Four physicians reviewed 12 responses from ChatGPT against two ASRM educational documents, utilizing a 1-5 rating scale. The findings revealed that while the AI's clarity was comparable to professional sources, it struggled with safety and accuracy, consistently overestimating or underestimating live birth rates.

The results were presented at the ASRM annual meeting in October 2026.

Culture Shift

This study highlights the growing friction between the rapid adoption of generative AI in healthcare and the rigid, high-stakes requirements of clinical accuracy. It reflects a broader societal challenge where convenient, conversational tools are outpacing established medical vetting processes.

Patients should exercise caution when using chatbots for medical advice and verify all fertility information against official medical resources. Relying on AI-generated data for life-altering reproductive decisions may result in receiving inaccurate or potentially unsafe guidance.

The takeaway

Patients should prioritize using accredited medical materials over AI chatbots when making sensitive decisions regarding fertility treatments. Verification of AI output with a qualified reproductive endocrinologist remains the safest path for patients seeking reliable health outcomes.