Artificial Intelligence for advancing Sustainable Development Goals a comprehensive thematic review

Artificial intelligence (AI) is increasingly recognized as a key enabler of environmental sustainability, with applications supporting climate action, resource efficiency, and progress toward the United Nations Sustainable Development Goals (SDGs). However, existing reviews are largely sector-specific and offer limited integrative insight into how AI simultaneously advances multiple SDGs through cross-sectoral synergies. Addressing this gap, this study examines AI applications across environment-related SDGs (SDG 2, 6, 7, 9, 11, 12, 13, and 15), spanning energy, agriculture and livestock, waste management, infrastructure, circular economy, and disaster preparedness. A narrative thematic review approach was adopted using structured searches across IEEE Xplore, ScienceDirect, SpringerNature Link, EBSCO, academic repositories and selected grey literature sources, including Google Scholar, SSRN, ResearchGate, and reports from the United Nations, IMF, and the Ellen MacArthur Foundation. A PRISMA-aligned screening process resulted in 180 sources for thematic analysis. The findings show that AI enhances climate modelling, enables precision agriculture, supports renewable energy integration, strengthens disaster preparedness, and contributes to sustainable urban and infrastructure systems. Sustainability outcomes are amplified when AI is integrated with complementary technologies such as the Internet of Things, blockchain, and remote sensing, particularly in advancing circular economy practices. Despite these benefits, challenges persist related to ethical governance, data privacy, algorithmic bias, digital inequality, and the environmental footprint of AI systems. The study extends the Technology–Organization–Environment and Diffusion of Innovation frameworks and offers policy-relevant insights for responsible and inclusive AI deployment, while outlining future research directions on low-resource AI, AI–SDG trade-offs, and impact measurement.