AI chatbot provides incorrect cancer treatment recommendations

[News]

Researchers evaluate the reliability of large language models (LLMs) in providing cancer treatment recommendations, finding that 33.33% of the chatbot's advice was partially non-concordant with established guidelines. The study highlights the need for caution and regulation as AI technologies infiltrate healthcare decision-making.

More... (https://www.news-medical.net/news/20230829/AI-chatbot-provides-incorrect-cancer-treatment-recommendations.aspx)