Artificial Intelligence and Sustainability: A Bibliometric Analysis and Future Research Directions

What the paper says

Background: The proliferation of research on artificial intelligence (AI) and sustainability has increased the lack of perspective on how future research can contribute to the big picture of sustainable development. This paper aims to synthesize and analyze academic research on AI and sustainability to reveal the main trends and propose a robust agenda to tackle future research on the theme. It answers four main research questions: (i) what is the current state of research on AI and sustainability? (ii) which are the most productive countries and journal outlets in this research area? (iii) how has the research in the area evolved? (iv) what are the research lacunae and, thus, the opportunity for future exploration? Method: To answer the research questions, we performed a bibliometric analysis of 3887 documents extracted from the Web of Science core collection of databases. Results: The primary finding of this research is that the motor themes pushing the research in AI for sustainability are related to energy efficiency, smart grid, and renewable energy. Yet the field suffers from eight main shortcomings: overreliance on ML; lack of study on human responses to climate crisis mitigation strategies; lack of performance measurement; lack of research about how cybersecurity risks may impact sustainable development efforts; lack of research about the adverse impact of AI development on the environment; lack of research on the impact of economics on AI for sustainability efforts; lack of discussion about policymaking and policy recommendation; and excessive focus on renewable energy. Conclusion: This paper contributes to scholarly conversations on the direction research on AI for sustainability should take by highlighting its shortcomings and proposing a robust research agenda to address them.