의료 인공지능의 활용과 의료 환경의 변화에 따른 법적 쟁점: 협업, 의무, 그리고 책

The healthcare landscape is undergoing fundamental changes with the commercialization of medical AI. First, the integration of healthcare and medicine, which has been underway since 2000, is expected to accelerate as structured and unstructured medical data are collected from various sources in everyday life. Data-driven medicine based on a comprehensive healthcare system will lead to precision and personalized medicine, shifting from a diagnosis-treatment paradigm centered on healthcare providers and medical institutions to a prevention-management paradigm centered on patients and consumers. The introduction of medical AI is expected to change the doctor-patient relationship, medical information asymmetry, and the doctor's duty to explain. Furthermore, in the new medical environment brought about by AI systems, the so-called three elements of medicine―the subject(physicians), the act(medical practice), and the means(medical devices)―should be reconstructed and reinterpreted. In AI systems such as Watson for Oncology(WFO), the subject and the means of medical practice are difficult to be clearly distinguished, and the importance of collaboration between physicians and medical AI is being increasingly emphasized. This can be evaluated in conjunction with the existing trends and tendencies of expanding multi-layered collaboration in the medical field, and meaningful reflection can be made on the stubborn regulation of unlicensed medical practice. Moreover, this collaborative medical practice is also essential in dealing with the issue of the attribution of responsibility for medical malpractice in the event of a medical accident that occurs when medical AI is used. This is a question of how the standard of the physician’s duty of care will change with the commercialization of medical AI. Here, the level of automated decision-making or the type of medical AI could be taken into consideration. In particular, the ‘principle of trust’ devised in the context of collaboration and division of labor among diverse personnel could be reviewed.