Artificial Intelligence for Sustainable Development: A Multidisciplinary Review of Applications, Challenges, and Future Research Directions

Abstract

Artificial intelligence is increasingly shaping sustainable development by improving prediction, optimisation, monitoring, and decision-making across environmental, social, agricultural, industrial, and economic systems. This review examines the multidisciplinary applications of AI in climate prediction, renewable-energy management, biodiversity conservation, precision agriculture, food systems, water-resource management, healthcare, education, smart cities, green manufacturing, and supply-chain resilience. It also evaluates the technical, ethical, social, environmental, and regulatory challenges that influence the responsible use of AI. The reviewed evidence shows that AI can strengthen resource efficiency, enhance service delivery, support risk assessment, and improve institutional responses to complex sustainability problems. At the same time, its benefits are constrained by poor data quality, limited infrastructure, algorithmic bias, privacy concerns, unequal access, weak governance, and the growing energy and carbon costs of large-scale computing. The review highlights the need to distinguish between using AI for sustainability and ensuring the sustainability of AI itself. Responsible governance, transparent models, human oversight, inclusive datasets, impact assessment, and cross-sector collaboration are essential for aligning AI with the Sustainable Development Goals. Future research should prioritise low-energy systems, explainable models, locally relevant data, and evidence from low- and middle-income regions. A multidisciplinary approach can help ensure that AI contributes to resilient, equitable, and environmentally responsible development.