A RESPONSABILIZAÇÃO CIVIL DAS BIG TECHS PELO TREINAMENTO DE I.A. À LUZ DA RACIONALIDADE PROTETIVA DO DIREITO AUTORAL BRASILEIRO E DA TEORIA DA CEGUEIRA DELIBERADA DA JUSTIÇA

Abstract

This study analyzes the civil liability of Big Tech companies for the use of copyright-protected data in the training of Artificial Intelligence (AI) systems, from the perspective of the Theory of Willful Blindness. The main objective of the study was to verify whether it is possible to establish the civil and criminal liability of major corporations for AI training by framing it within the protective rationale of Brazilian copyright law and the theory of willful blindness in justice. The study adopted an inductive approach method, through bibliographic research and qualitative documentary analysis, with an emphasis on constitutional and infra constitutional interpretation, as well as doctrinal, normative, and jurisprudential analysis. To achieve this objective, the study is divided into four specific stages. The first explores the advancement of Artificial Intelligence and its dependence on data in the context of large technology corporations. It includes an analysis of copyright law within the current Brazilian legal framework and examines how the use of protected content in AI training processes may violate national legislation. It also discusses how companies such as Meta, Google, and OpenAI benefit from the absence of specific regulation to avoid sanctions, often claiming ignorance regarding the illicit origin of the data. The second stage investigates the Theory of Willful Blindness through doctrinal and jurisprudential research, highlighting its emergence, legal foundations, and contextualization, and seeks to frame it as a basis for holding Big Tech companies accountable for their systematic copyright violations. The third stage offers an integrated analysis of different branches of law — civil, criminal, and corporate — focusing on liability within each legal sphere, addressing its evolution and application in both the civil and criminal domains, as well as in corporate law. The research discusses the application of strict civil liability (Article 927, Civil Code/2002) and the criminal classification of offenses (Article 184, Penal Code), emphasizing both international and national precedents. The fourth stage argues that the willful omission by Big Techs and regulatory inertia constitute systematic violations, requiring a robust legal framework inspired by the European AI Act to ensure transparency, data traceability, and compliance.