Título: Ethical Challenges in Artificial Intelligence: Navigating the Boundaries of Responsible Innovation
Autores: Julio Cesar Sousa de Lima, Felipe Oliveira e Silva & Danton Diego Ferreira
Resumo: The integration of Artificial Intelligence (AI) in Industry 4.0 and 5.0 demands robust ethical frameworks; however, the practical application of principles such as fairness, accountability, and transparency in Computational Intelligence (CI) techniques remains a challenge. This study presents a Systematic Literature Review (SLR), following PRISMA (Preferred Reporting Items for Systematic reviews and Meta-Analyses) guidelines, analyzing peer-reviewed articles (from 2018 to early 2025) from six databases on how these principles are addressed. The review maps the discourse across eight thematic axes, including algorithmic bias mitigation, explainability (XAI), human-centric AI design, and governance mechanisms. Findings reveal a persistent gap between widely accepted ethical principles and their fragmented implementation in socio-technical systems. Although human-centered approaches and fairness-aware tools continue to advance, effectively operationalizing ethics still requires further development. This review consolidates the current state of research, identifies key challenges facing the CI community, and outlines directions for promoting ethically aligned AI innovation in industrial contexts.
Palavras-chave: Artificial Intelligence; Ethics; Industry 4.0; Human-Centric AI; Fairness; Accountability; Governance; Algorithmic Bias; Systematic Literature Review.
Páginas: 8
Código DOI: 10.21528/CBIC2025-1143995
Artigo em PDF: CBIC_2025_paper1143995.pdf
Arquivo BibTeX:
CBIC_2025_1143995.bib