Medical damage liability risk of medical AI: from the perspective of DeepSeek’s large-scale deployment in Chinese hospitals

The field of healthcare is one of the important areas for the application of artificial intelligence (AI). This study introduces the current deployment of the AI model DeepSeek in Chinese hospitals, raises concerns about the ethical and legal aspects of medical AI, and identifies the problem of insufficient regulation by reviewing the current regulatory status of medical AI in China. In the discussion section, this article mainly focuses on three types of medical damage liability risks in medical AI, namely medical product liability, diagnosis and treatment damage liability, and medical ethics liability. In the determination of medical product liability, the ethical attributes and technological characteristics of medical AI determine its auxiliary positioning, but the auxiliary positioning of medical AI has not eliminated the applicable space of medical product liability, and in the judgment of product defects, the "rational algorithm" standard based on the "rational person" standard should be used to identify AI design defects; In the determination of diagnosis and treatment damage liability, medical AI has not changed the existing doctor-patient relationship structure, but the human-machine collaborative diagnosis and treatment model has intensified the difficulty of identifying doctor's fault, so "reasonable doctor" standards should be adopted, and medical personnel should be given the discretion to reevaluate the negligence of doctors in using AI recommendations. In the case of localizing DeepSeek deployment in hospitals, if misdiagnosis occurs, hospitals and doctors are more likely to bear the diagnosis and treatment damage liability rather than medical product liability. At the same time, the adoption of DeepSeek exacerbates the lack of protection for patients' right to informed consent, which may lead to medical ethical liability. In addition, this article also discusses the data compliance risks of large-scale deployment of DeepSeek in hospitals.