LONDON (Kashmir English): A new study has warned that people should not fully rely on artificial intelligence (AI) tools such as ChatGPT for medical advice as it could ruin their health.
The study has warned that people should exercise caution when seeking health-related guidance from AI platforms.
According to a report, more than 230 million people worldwide turn to ChatGPT each week with health-related questions.
Many ask the platform about symptoms of illnesses, possible home remedies for various medical conditions or safe foods.
The research found that while the AI tool can often correctly recognize clear medical emergencies, it underestimated the seriousness of more than half of the cases where immediate medical attention was required.
For the study, researchers created 60 different medical scenarios covering 21 medical specialties. These scenarios ranged from minor health issues to severe emergency situations.
AI response less reliable for medical advice
The findings showed that the AI’s responses were less reliable when users asked questions about sensitive issues such as self-harm.
In such cases, researchers noted that the responses sometimes appeared inconsistent or contradictory.
One of the study’s co-authors, however, emphasised that the findings do not mean AI should be completely avoided in healthcare. Instead, the technology can still be helpful if used carefully and alongside professional medical guidance.
Experts further advised that individuals experiencing symptoms such as severe allergic reactions, chest pain, or rapidly worsening conditions should seek immediate medical attention rather than relying solely on AI advice.
The study also highlighted that AI language models are continuously being updated and improved, meaning their performance may change over time.
The study found that there are risks to people seeking medical advice from AI due to their tendency to provide inaccurate and inconsistent information. The results have been published in Nature Medicine.
