Bridging Data Protection and AI Ethics: A Two-Study Empirical Examination of LGPD Principles and Ethical AI in Brazil

Context: Data protection laws and AI ethics frameworks are increasingly invoked to govern the risks of data-driven and algorithmic systems, but there is still limited empirical evidence on how practitioners and students perceive the relationship between legal principles (such as the Brazilian LGPD) and ethical principles for Artificial Intelligence (AI). Goal: This study investigates how LGPD principles are interpreted as a foundation for ethical AI, examining perceived alignments, practical challenges, regulatory gaps, and expectations for the evolution of AI governance in Brazil. Method: We conducted two complementary survey based studies. Study 1 collected responses from 30 computing students, exploring their perceptions of privacy, transparency, security, data minimization, and accountability in AI systems. Study 2 extended this investigation with 100 participants (students and professionals in diverse software project roles), using paired LGPD AI items, Likert-scale questions on sufficiency and complementarity, and open-ended questions analyzed through inductive content analysis. Results: Across both studies, participants consistently perceived strong conceptual alignment between LGPD principles and AI ethical principles, especially regarding privacy, transparency, security, prevention, non-discrimination, and accountability. However, they also reported important gaps, particularly in explainability, fairness and bias mitigation, inclusion and diversity, solidarity, and other human-centered values, as well as uncertainty about accountability for automated decisions and the fit of static legal principles to dynamic AI systems. Conclusion: The findings indicate that LGPD is viewed as a solid but insufficient foundation for ethical AI. Participants expect LGPD to be complemented or updated by AI specific ethical and regulatory frameworks, governance mechanisms, and technical measures such as explainable AI and algorithmic auditing, pointing to the need for an integrated ecosystem for responsible AI in Brazil.