Artificial intelligence for environmental sustainability: opportunities, ethical challenges and future research directions

Purpose This paper aims to examine the application of artificial intelligence (AI) to environmental sustainability, with particular attention to ethical and governance factors that shape whether AI contributes to meaningful sustainability outcomes. Design/methodology/approach This paper conducts a structured qualitative literature review of peer-reviewed studies on the applications of AI to environmental sustainability, with particular attention to ethical and governance concerns and emerging research directions. Findings The analysis shows that AI is most frequently applied in climate modelling and forecasting, energy optimization, waste management, biodiversity conservation and water-resource management. Across these domains, AI is used primarily to support analytical functions such as prediction, monitoring, optimization and model-based decision-making. Nevertheless, AI does not automatically generate sustainability benefits. Its contribution is conditioned by factors such as the energy demands of AI systems, infrastructure availability, governance capacity and ethical concerns relating to transparency, fairness and accountability. Practical implications The article provides practical policy recommendations for policymakers and emphasizes the importance of incorporating sustainability criteria into the governance of AI, enhancing the transparency and accountability obligations for climate-related AI solutions, promoting the development of “green AI” solutions that focus on energy efficiency and supporting capacity development and data infrastructure. A short-term, medium-term and long-term research agenda on the measurement and evaluation of AI is proposed for more comprehensive integration. Originality/value This review contributes to the literature in three ways. First, it integrates application-focused and ethics/governance-focused studies on AI and environmental sustainability within a single synthesis. Second, it argues that AI should be understood as a conditional enabler of sustainability rather than as an inherently sustainable solution. Third, it develops a conceptual synthesis model showing how AI applications generate sustainability-relevant analytical capabilities, but how the translation of those capabilities into sustainability outcomes depends on governance quality, ethical safeguards and institutional capacity.