The contemporary world scenario demands initiatives to address current societal challenges. Among them, the Environmental, Social, and Governance (ESG) criteria have received prominent attention. Traditionally, ESG is adopted to handle investments. Therefore, by adopting interdisciplinary and integrative approaches, ESG has the potential to bring insights, measurements, understandings, and evaluations on how companies and their economic processes affect society in several layers, from a local to regional and global view. This study presents a comprehensive bibliometric analysis of 55,000 publications supported by Natural Language Processing (NLP) techniques on the aspects of ESG-related topics from a systemic standpoint. The ESG theme has been studied extensively, following a linear trend. However, even with this increasing attention, it is essential to remark that there are challenges to data availability that conducts a non-balanced context between environmental, social, and governance aspects. Concerning the E criteria, besides the increasing initiatives related to energy management and ghg emissions, there are several challenges to overcome, for example, the lack of investments, regulation, and inspection in most countries, especially in developing economies. In the S context, labor practices and employee health and safety issues were found. It may result in violations of fundamental human rights, freedom of association, and a safe, physical and mental environment. With respect to G, business ethics has demonstrated failure, meaning problems with moral norms, accountability, transparency, justice, and rights. Furthermore, by adopting the General System Theory, it is possible to characterize ESG in terms of its connected parts, interaction with external medium, equifinality, feedback, and anamorphism. Such analysis brings valuable information, serving as a support tool in decision-making processes involving ESG, thus, promoting a more sustainable society and world
