Human-Centered AI to Accelerate the SDGs: Evidence Map (2020–2024)

Artificial Intelligence (AI) has gained prominence on sustainability agendas while raising ethical, social, and environmental challenges. This study synthesizes evidence and maps the scientific production on Human-Centered AI (HCAI) at the interface with the Sustainable Development Goals (SDGs) for 2020–2024. Searches in Scopus and Web of Science (Boolean operators; thematic and temporal filters), followed by deduplication, yielded 265 articles, which were analyzed with Bibliometrix/Biblioshiny version 5.1.1 and VOSviewer version 1.6.20 (0) to generate term co-occurrence maps, collaboration networks, and bibliographic coupling. The results indicate accelerated growth and diffusion of the topic, with journals such as Sustainability, IEEE Access, and Applied Sciences standing out. Three interdependent axes were identified: (i) technical performance, with emphasis on machine learning and deep learning; (ii) explainability and human-centeredness (XAI, ethics, and algorithmic governance); and (iii) socio-environmental applications oriented toward the SDGs. Underrepresentation of the Global South, particularly Brazil, was observed. It is concluded that HCAI is being consolidated as an emerging interdisciplinary field with potential to accelerate the SDGs, although there remains a need to integrate ethical, regional, and impact-assessment dimensions more systematically to achieve global targets effectively.