Study Warns of Inaccuracies in Chatbot-Provided Medication Information

Researchers highlight potential risks of relying on AI-generated drug advice due to frequent inaccuracies and complexity.

Researchers from the University of Erlangen conducted a study on the reliability of AI chatbots for medication information.
The study, published in 'BMJ Quality & Safety,' found that chatbot responses were often inaccurate and difficult for non-experts to understand.
Researchers evaluated 500 AI-generated responses to common questions about 50 widely prescribed medications in the U.S.
Despite improvements in AI technology, the study emphasizes that chatbots still pose risks to patient safety due to incomplete or incorrect information.
The researchers call for clear warnings that AI-generated information should not replace professional medical advice.