In New York, a China-based AI healthcare technology company named Future Doctor, in collaboration with 32 clinical experts, has released a new study in Nature Portfolio’s npj Digital Medicine presenting a unique assessment method called the “Clinical Safety-Effectiveness Dual-Track Benchmark” (CSEDB). This framework aims to evaluate the safety and effectiveness of medical AI systems in practical clinical scenarios. The research includes a comparative analysis of prominent large language models like OpenAI’s o3 and Google’s Gemini 2.5 P.