Abstract:
The increasing development of artificial intelligence (AI) and its incorporation into
essential activities such as healthcare, transportation, finance, and public
administration has provoked profound reflections in the legal field, especially regarding
civil liability. The autonomy of these systems, capable of making decisions without
direct human intervention, challenges traditional legal concepts, since Brazilian law still
lacks specific normative instruments to address liability for damages arising from their
use. This study aims to analyze who should be held liable for damages caused by
autonomous AI systems in Brazil, considering normative gaps, limitations of current
legislation, and international experiences. To this end, a qualitative approach is
adopted, with bibliographic and documentary review, based on legal doctrine, national
and foreign legislation, and regulatory initiatives. The results show that traditional
models of civil liability, based on fault or the theory of risk, are insufficient in the face of
the autonomous logic of AI, requiring the construction of new legal frameworks capable
of ensuring legal certainty and effective protection for victims. Thus, the study points to
the need for Brazil to advance in the formulation of specific norms, possibly based on
objective responsibility and mechanisms of shared accountability among agents that
are part of the technological chain.
