A new survey suggests that Gen Z is turning to AI chatbots for sexual health advice at a pace that would have been unthinkable just a few years ago. According to data released by Testing.com, about 1 in 10 Gen Z adults in the U.S. have asked ChatGPT or another AI chatbot to help diagnose a possible sexually transmitted disease. Even more striking, many of them are sharing intimate details and photos in the process, despite knowing the privacy risks.
The survey, conducted in January 2026, polled 2,520 U.S. adults between the ages of 18 and 29. It found that 41 percent of Gen Z respondents have used an AI chatbot for medical advice. Of that group, about half said they specifically asked about STIs or STDs, and roughly half of those were seeking an actual diagnosis rather than general information. That means around 20 percent of Gen Z has asked AI about STDs, and about 10 percent has tried to get a diagnosis from a chatbot instead of a clinician.
What Gen Z is asking about is revealing. Among those who turned to AI for STI and STD questions, 60 percent asked about symptoms, 59 percent asked about prevention, and 57 percent asked about possible exposure. Nearly half asked about treatment, and 45 percent asked directly about diagnosis. A smaller but still notable group asked about testing and even how to talk to a partner about a possible infection.
To get answers, many users went far beyond vague questions. Ninety two percent shared a list of symptoms, 76 percent shared details about recent sexual activity or possible exposure, and more than half shared photos. Forty four percent even shared a partner’s known STI or STD status. Based on that information, 45 percent say the chatbot indicated they had an infection, 36 percent were told they did not, and 17 percent got an uncertain answer.
Doctors say that is where the real danger begins.
“An AI chatbot can be a starting point for general education, but it shouldn’t be used as a diagnostic tool for an STI,” says Toni Brayer, M.D., an internal medicine doctor and member of Testing.com’s medical review board. “The information provided is dependent upon the question asked, and many symptoms can be caused by non-STD conditions. A chatbot is not a replacement for medical professionals or actual testing (blood tests, urine tests, swabs, and cultures), which is still the only way to confirm infection. If a chatbot leads someone to delay testing, delay treatment, or avoid notifying a partner, that’s where the real harm happens.”
The numbers back up that concern. After their diagnostic conversation with ChatGPT, 74 percent of users said they went on to get tested, but 26 percent did not. Among those who got tested, 31 percent tested positive and 68 percent tested negative. Overall, AI misdiagnosed 31 percent of users.
The errors went both ways. Among users who were told by AI that they did not have an STI or STD, 15 percent later tested positive. Among those who were told they did have one, 53 percent later tested negative.
“AI chatbots should be used only to gather information, and they are not diagnostic tools,” Dr. Brayer explains. “STIs require lab testing to diagnose accurately. A false negative or a false positive rate of 31 percent is unacceptable in medical diagnosing. When AI gets it wrong, it can cut both ways: someone may feel falsely reassured and delay care when treatment is needed, or they may be told they have an STI when they don’t, leading to unnecessary anxiety and stress.”
Privacy is another major concern, and Gen Z knows it. Nearly a third of respondents said they were moderately concerned about sharing medical information with AI chatbots, while 20 percent said they were very concerned and 13 percent said they were extremely concerned. Even so, future use remains high. Ninety two percent say they are very or somewhat likely to talk with an AI chatbot again about STIs or STDs.
Interestingly, comfort levels are split. Thirty five percent say they are more comfortable talking to an AI chatbot than a medical professional about STDs, while 25 percent prefer a doctor. Another 36 percent say they are equally comfortable with both.
“When you share sexual health details or photos with an AI chatbot, you don’t have the same privacy protections you’d have in a medical setting under HIPAA,” Dr. Brayer warns. “Depending on the platform, information could be stored or used to improve the system. Your personal data can be compromised, sold to vendors, or otherwise used in ways that are harmful. The legal protections are not in place for confidential medical information.”
The takeaway is not that Gen Z is reckless, but that it is pragmatic, anxious, and deeply comfortable with technology. AI is fast, anonymous, and always available. For a generation that grew up online, it feels natural to ask a chatbot first. The problem is that bodies are not search queries, and medicine still requires tests, professionals, and real privacy protections.
For now, some doctors say AI can be useful for education, but warn diagnosis should stay in the clinic, not the chat window.
