医 疗 人 工 智 能 的 前 端 准 入 规 制 与 中 后 端 侵 权 责 任 调 适

Based on the unique characteristics of medical artificial intelligence（AI）that differ from general medical devices，it was found that the clinical application of medical AI has brought different levels of impacts on the front-end access supervision and regulation，the mid-end informed consent right design，and back-end post-damage liability adjustment. In this regard，it is necessary to appropriately relax the admission review，strengthen the coordination of supervision before and after listing，and simultaneously adjust the comprehensive judgment ideas of“safe and effective”medical AI；clarify the“breadth”of notification and“depth”of disclosure regarding the patients’informed consent rights；based on current product liability and liability for medical malpractice system， typologically adjust the allocation of liability， while strengthening the supplemental application of medical compulsory liability insurance，so as to enhance the relief of patients’damage，rationalize risk allocation，and minimize social costs.