At present artificial intelligence AI technology is rapidly expanding in the medical field and the issue of liability identificationin artificial intelligence-assisted medical treatment is becoming increasingly prominent. Due to the characteristics of medicalartificial intelligence itself and the lack of relevant laws in China the identification of liability for patient harm caused by man-machinecooperation medical activities still blindly adopts the traditional liability for medical malpractice and the true liability subject has notbeen effectively regulated. Through comparative research methods and case analysis this paper studies the specific attribution issuesof AI-assisted medical malpractice discusses the role positioning and obligations of human physicians in the man-machine cooperationsystem and concludes that the product attribute and auxiliary status of medical artificial intelligence should be adhered to when the infringementaccident of AI-assisted medical treatment occurs. By reducing the difficulty of proof of the victim and expanding the scopeof the subject of the producer the rules of liability for medical damage and medical product liability should be properly adjusted toclarify the subject of liability and the way of liability under different circumstances to protect the rights and interests of all parties.