Sustainable generative AI and quantum computing: review assessment on the environmental impact of generative AI and quantum technologies

The rapid advancement of Generative Artificial Intelligence (GenAI) and Quantum Computing (QC) presents transformative opportunities, yet their high computational requirements raise concerns about their environmental sustainability. This comprehensive review examines the ecological footprint of both technologies, focusing on key metrics like energy consumption, carbon emissions, and resource depletion. Findings from existing studies consistently indicate that the impact of GenAI is mostly driven by the immense energy demands of large-scale model training and inference. Moreover, findings from the review reveal that the footprint of QC largely stems from the energy-intensive cryogenic cooling and rare material requirements of its specialized hardware. This paper benchmarks current approaches to environmental assessment, highlighting the important role of Life Cycle Assessment (LCA) in providing a holistic view of the classification of environmental impacts across the entire supply chain, from manufacturing to disposal. This study proposes a range of domain-specific mitigation strategies, including algorithmic optimizations like pruning and distillation for AI, and cryogenic and material sourcing improvements for quantum systems. This study also proposes a framework for proactive, responsible innovation and identifies some gaps in the literature, such as the lack of standardized metrics and transparent reporting. There is a need to embed eco-conscious principles in the design of future technologies and highlight opportunities where these technologies can be used to handle broader climate challenges. The findings in this study can be used by policymakers, researchers, and industry stakeholders in aligning technological progress with global climate and sustainability goals.