Chatbots remain imperfect and can sometimes give inaccurate or even harmful information. A study by European scientists found that in 22% of cases, medical advice from AI systems like Copilot could lead to serious harm, including fatal outcomes.
The research analyzed 500 responses to common medication-related questions. Of these, 24% were inaccurate, and 3% were entirely incorrect. Only 54% aligned with evidence-based medical principles, demonstrating a significant gap in reliability.
Potential Health Impact of Incorrect Responses
The study revealed that 42% of the bot’s responses could cause mild to moderate harm, while 22% could pose serious risks, including life-threatening situations. Only 36% of the answers were deemed safe, underscoring the need for caution when using AI for medical inquiries, notes NIXSOLUTIONS.
Experts Recommend Consulting Professionals
Scientists strongly advise against relying solely on chatbots for medical advice. Instead, people seeking accurate health information should consult qualified healthcare professionals. Relying on human expertise ensures a higher standard of care and reduces the risk of harm.
We’ll keep you updated on developments in AI safety to help you make informed decisions about your health.