Integrating Artificial Intelligence and Climate Sustainability for Human Health: A Systematic Review of Natural Resource Challenges, Healthcare Opportunities, and Policy Directions

Abstract

The field of artificial intelligence (AI) and climate sustainability, climate protection and natural resources management, and health science has become an interdisciplinary research area in which the digital transformation and climate protection are to be assessed together. AI has become a more common feature in medical imaging, aids in clinical decision making, disease surveillance, predictive analytics, drug discovery, health-system management, telemedicine, and population-health monitoring. Such applications can enhance diagnostic efficiency, ensure optimal resource allocation, minimize unnecessary healthcare utilization, reinforce climate-sensitive disease monitoring, and aid in resilient healthcare systems. Meanwhile, AI systems also need significant amounts of computing power, electricity, cooling water, semiconductors, data storage, networking and electronics. This has led to the environmental impact of AI's use being a pertinent issue for sustainable healthcare.
This systematic review explores the connection between AI and climate sustainability in human health outcomes, focusing on natural-resource use, health opportunities, environmental threats, health-system resilience and policy needs. Literature was synthesized in five analytical domains: AI-enabled health applications, climate-sensitive health risks, computational energy and carbon impacts, natural-resource impacts, and governance mechanisms. Healthcare systems have been estimated to be responsible for about 5% of global GHG emissions and data centres are estimated to be responsible for about 180 Mt of indirect CO2 emissions due to electricity consumption. While this is just a fraction of overall data centre workloads, AI's increasing appetite for compute power has contributed to an increase in strain on electric power systems, water resources, hardware supply chains, and electronic-waste management.
The review points to a two-fold path. Predictive maintenance, energy optimization, precision medicine, telemedicine, optimization of supply chains, prediction of diseases related to climate change, and better resource utilization are just a few examples of how AI can lessen environmental burdens when developed correctly.AI can ease environmental burdens in a variety of ways when designed appropriately, such as: predictive maintenance, energy optimization, precision medicine, telemedicine, optimization of supply chains, prediction of diseases related to climate change, and better resource utilization. A key component of sustainable deployment is lifecycle assessment, along with energy- and carbon-aware computing, renewable electricity procurement, water-efficient data-centre design, hardware circularity, transparent environmental reporting and algorithmic efficiency, plus health-specific governance. It is proposed to develop a policy framework that integrates AI governance with the policy of climate and health. Assessing the performance of sustainable AI for health should not be based solely on clinical performance but also on human-health, environmental, economic, equity, and technological performance.