Abstract
This paper explores the synergistic relationships among responsible artificial intelligence (RAI), environmental, social and governance (ESG), and sustainable development goals (SDGs), highlighting how their integration can create opportunities for organizations to enhance societal impact while increasing corporate value and fostering innovation. Through a systematic literature review, this study first provides foundational knowledge related to RAI, ESG, and SDGs, including their origins, core concepts, and main implementation challenges. Then, the relationships between RAI, ESG, and SDGs are examined and discussed through the analysis of connections and synergies among them. Based on this analysis, an integrated framework that synthesizes the relationships among RAI, ESG, and SDGs is proposed, demonstrating how they can be mutually reinforcing when implemented cohesively. Additionally, the paper discusses implications of this integrated approach for multiple stakeholders, while acknowledging implementation challenges and tensions. The study emphasizes the necessity of viewing RAI, ESG, and SDGs as interconnected frameworks rather than individual goals. Through this integrated lens, organizations can align RAI, ESG, and SDGs collectively in their strategic planning and operational practices, thereby promoting more effective and sustainable outcomes while addressing complex societal challenges in the AI era. The research contributes to both theoretical understanding and practical guidance for implementing holistic approaches that harness the transformative potential of AI technologies while ensuring that their development and deployment serve broader sustainability and social responsibility objectives.
| Original language | English |
|---|---|
| Pages (from-to) | 20-41 |
| Number of pages | 22 |
| Journal | IEEE Computational Intelligence Magazine |
| Volume | 20 |
| Issue number | 4 |
| Early online date | 10 Oct 2025 |
| DOIs | |
| Publication status | Published - 1 Nov 2025 |
Bibliographical note
Publisher Copyright:© 2005-2012 IEEE.
Funding
This work was supported in part by the National Natural Science Foundation of China under Grant 62250710682, in part by the National Key R&D Program of China under Grant 2023YFE0106300, in part by Guangdong Provincial Key Laboratory under Grant 2020B121201001, and in part by the Program for Guangdong Introducing Innovative and Entrepreneurial Teams under Grant 2017ZT07X386.