Benefits
Large AI companies such as OpenAI and Anthropic are developing new features that allow chatbots to answer questions related to health and health.
According to Robert Wachter, a medical technology expert at the University of California San Francisco (USA), AI tools are capable of providing more specific and personalized information.
Sometimes there are no alternatives, or patients have to find their own way to find information. Therefore, if used responsibly, these tools can still provide useful information," said Mr. Wachter.
Another advantage of the new generation chatbots is the ability to answer based on personal data such as age, medical history, prescriptions or doctor's notes.
Even without allowing access to medical records, experts still encourage users to provide as much information as possible for chatbots to provide more accurate feedback.
When should I not ask AI
However, doctors warn that chatbots are not suitable in emergency situations. Symptoms such as shortness of breath, chest pain or severe headache can be signs of medical emergency and need to be checked immediately at the hospital.
Lloyd Minor, president of Stanford University School of Medicine (California, USA), believes that users should access AI tools with "a healthy level of skepticism".
According to him, even if it is just a small health decision, users should not just rely on information from chatbots to make decisions.
Consider privacy issues
Another important factor is health data security. Many benefits of chatbots come from users sharing personal health information. However, these data are not protected according to the same standards as medical records in the hospital system.
In the US, the HIPAA health security law applies to doctors, hospitals and insurance companies, but does not apply to AI chatbot development companies. Therefore, users need to consider carefully before uploading personal health records.
According to Adam Mahdi, a researcher at the Oxford Internet Institute (under the University of Oxford), chatbots can accurately identify the condition of the disease in many theoretical scenarios, but problems may arise when interacting with actual users.
Many users do not provide enough necessary information, making it difficult for AI to give accurate assessments. Meanwhile, chatbots sometimes respond by mixing in both correct and inaccurate information, making it difficult for users to distinguish which content is reliable.
Some experts believe that the best way to use chatbots is to see them as a supplementary reference source, similar to asking for a second opinion from a doctor.
According to Mr. Wachter, sometimes he enters the same question into ChatGPT and other AI tools such as Google's Gemini. If chatbots give similar results, he will feel more confident about the information received.
