Abstract: This study investigates the medical ethics cognition characteristics of medical school students in the era of artificial intelligence through word frequency analysis and empirical surveys. The results show that students' cognition presents a risk-cause-prospect three-dimensional structure: the risk dimension focuses on data privacy and security,the cause dimension concerns algorithm fairness and ambiguous liability attribution,and the prospect dimension looks forward to improved medical efficiency and the ethical education system.Professional category is the core variable of cognitive differences,with students majoring in clinical medicine and medical technology having higher cognitive levels than students in intelligent medicine majors.Gender,grade level,and clinical internship duration have no significant impact.Students' cognition shows limitations with an overly strong perception of risk and weak awareness of policy governance.The study proposes strategies such as building a legal framework and strengthening privacy protection to provide references for balancing the efficacy of AI in healthcare and ethical risks.