A morte da responsabilidade civil médica: uma análise da inteligência artificial na medicina diagnóstica

Abstract

Artificial intelligence (AI) has been an important tool for technological development, directly driving the speed, breadth, and evolutionary depth of new technologies. The exponential nature of this development stems from increased data processing capacity combined with its widespread availability. Its influence has had a significant impact on the social, economic, cultural, and business architecture of humanity, especially in decision-making. The complexity of a physician's civil liability, which already required intense reflection on the interaction of variables that directly or indirectly interfere with their activity, such as violation of medical conduct duties, therapeutic risk, precarious infrastructure, and educational deficiencies, adds new layers with the inclusion of AI in this scenario. Within diagnostic medicine, the sensitivity is even greater, because while it achieves levels of diagnostic accuracy superior to human capabilities with direct benefits to the patient and the healthcare system itself, it faces serious structural problems, such as the existence of discriminatory algorithms, opacity, lack of transparency, limited database training, lack of interoperability, and risk to personal data protection. This research will therefore focus on this new scenario in health and medicine, specifically on the use of AI in diagnostic medicine, correlating its use and medical civil liability, specifically the analysis of the scope of the physician's civil liability within a factual-legal scenario of diagnostic medicine, when the AI system as a decision support fails, with an emphasis on software as medical devices or Software as a Medical Device (SaMD). The starting point is the premise that there is already unfavorable hyperbole regarding the performance of physicians, with the attribution of civil liability in non-compliance with the civil liability system adopted in the legal system, a fact that will be aggravated with the spread of AI with the capacity for decision-making in more effective and accurate diagnoses, requiring a revisiting of the civil legal institutions of civil liability, without prejudice to the more fruitful action of the Legislative Branch and regulatory bodies, such as ANVISA, the Ministry of Health, and the Federal Council of Medicine, to reestablish a stable dogmatic system. This study aims to contribute to the reflection on the path that has been taken by the implementation of AI in health and diagnostic medicine and the factual and legal distortion of medical civil liability, seeking to establish subsidies for a better assessment of the attribution of civil liability.