How should you interpret AI responses? What should you watch out for?

AI loves to use medical terminology, but that doesn’t mean it has any medical knowledge. It’s still just a computer programme that strings words together into professional-sounding sentences, but doesn’t understand their meaning or the effects they might have on our bodies. Advice generated by artificial intelligence can be helpful – it can reassure us when we’re worried about test results, or suggest what questions to ask during a clinic visit.

However, we would like to draw your attention to a few key warning signs that should put us on our guard:

• DO NOT trust advice where the AI makes a diagnosis more definitively than it recommends a visit to the doctor.
• AI does not ask for a full medical history, allergies or other medicines you are taking (which a doctor will always do). In this case, on the one hand, we may feel a sense of security – it does not ask us for our personal details. On the other hand, we must realise that without a more detailed analysis of our health, AI is unable to select the appropriate treatment for us – only a doctor can do that!
• You should absolutely NOT trust information regarding specific dosages (e.g. of a medicine, supplement or herb) or claims that a particular supplement will ‘replace’ a medicine.
• If you wish to use an image (e.g. of skin symptoms or a CT scan) during a consultation with AI, bear in mind that it can only provide information based on similar images it finds online. Remember, however, that every case is unique! You must absolutely NOT blindly trust AI’s results.

Remember! Artificial intelligence (such as ChatGPT, Gemini or Google Search) is merely a computer programme. AI will never replace the professional expertise of a doctor, pharmacist or diagnostician!