A new white paper released by NTT DATA, titled “Sustainable AI for a Greener Tomorrow,” has sparked significant discussion regarding the environmental implications of artificial intelligence (AI). As AI technology continues to advance at an unprecedented pace, the report highlights the urgent need for sustainable practices in its development and deployment. The findings underscore a critical reality: the rapid growth of AI is placing unsustainable demands on the planet’s resources, necessitating immediate action from organizations and stakeholders across the industry.
The white paper was unveiled simultaneously in Tokyo and London on October 29, 2025, and it presents a comprehensive analysis of the environmental challenges posed by AI. Researchers predict that by 2028, AI workloads will account for more than 50% of data center power consumption. This staggering statistic raises alarms about the energy demands associated with training large language models, running inference pipelines, and maintaining always-on services. The implications of such energy consumption are profound, not only for the tech industry but also for global efforts to combat climate change.
One of the most pressing concerns highlighted in the report is the increasing electricity consumption required to support AI technologies. The computational needs of AI systems are immense, leading to a surge in energy use that could overwhelm existing infrastructure if left unchecked. This demand is compounded by the growing water consumption necessary for cooling systems in data centers, as well as the generation of electronic waste (e-waste) and the extraction of rare-earth minerals essential for hardware production.
David Costa, head of sustainability innovation headquarters at NTT DATA, emphasized the dual nature of AI’s impact on the environment. While the resource consequences of AI’s rapid growth are daunting, he noted that the technology also holds the potential to empower innovative solutions to the very environmental problems it creates. For instance, AI can be harnessed to manage energy grids more efficiently, model environmental risks, and improve water conservation efforts. However, for these benefits to materialize, organizations must recognize the challenge and prioritize sustainability in their AI systems from the outset.
The white paper advocates for a paradigm shift in how organizations approach AI development. It urges companies to move beyond traditional performance metrics such as accuracy and speed, which have dominated the conversation around AI effectiveness. Instead, the report calls for the incorporation of efficiency and sustainability as core design principles. This shift is crucial for ensuring that AI technologies contribute positively to environmental goals rather than exacerbating existing issues.
To facilitate this transition, the paper proposes the establishment of standardized and verifiable metrics to quantify AI’s environmental impact. Metrics such as the ‘AI Energy Score’ and ‘Software Carbon Intensity for AI’ would provide organizations with benchmarks to assess their energy use, carbon emissions, and water footprint. By adopting these metrics, companies can better understand their environmental footprint and take actionable steps toward reducing it.
NTT DATA’s researchers advocate for a lifecycle-centric approach to AI, emphasizing the importance of incorporating sustainability considerations at every stage—from raw material extraction and hardware manufacturing to system deployment and eventual disposal. This holistic perspective is essential for addressing the multifaceted challenges posed by AI’s resource demands. For example, extending hardware lifespans, optimizing cooling systems, and adopting circular economy principles can significantly mitigate the environmental impact of AI technologies.
Despite the growing awareness of sustainability issues, many organizations continue to focus narrowly on energy use or carbon emissions without considering other critical factors such as water consumption, depletion of rare materials, and e-waste generation. Even when environmental targets are established, few organizations have clear methods to integrate sustainability throughout the AI lifecycle. This gap in understanding and action underscores the need for comprehensive strategies that encompass all aspects of AI development and deployment.
To address these gaps, the white paper recommends several actionable strategies. One key suggestion is to apply green software engineering patterns to reduce resource consumption. By designing software that is inherently more efficient, organizations can minimize the energy and resources required to run AI workloads. Additionally, the report encourages organizations to run AI workloads in locations and at times that coincide with renewable energy availability. This practice not only reduces reliance on fossil fuels but also aligns AI operations with broader sustainability goals.
Another recommendation involves leveraging remote GPU services, which can help distribute computational loads more efficiently and reduce the need for on-premises hardware. By prioritizing modular, upgradable components, organizations can also reduce e-waste and extend the lifespan of their hardware. Refurbishing or recycling existing components further contributes to a more sustainable approach to AI infrastructure.
As the demand for AI continues to grow, so too does the responsibility of organizations to build these technologies responsibly. The white paper serves as a clarion call for the tech industry to embrace sustainability as a fundamental principle in AI development. By recognizing the environmental challenges posed by AI and taking proactive steps to address them, organizations can ensure that they contribute to a greener future.
In conclusion, the NTT DATA white paper highlights the urgent need for sustainable practices in the rapidly evolving field of artificial intelligence. As AI workloads are projected to consume a significant portion of data center power in the coming years, it is imperative for organizations to embed sustainability into every stage of AI development and deployment. By adopting a lifecycle-centric approach, establishing standardized metrics, and implementing green engineering practices, the tech industry can harness the potential of AI while minimizing its environmental impact. The path forward requires collaboration, innovation, and a commitment to responsible AI that prioritizes the health of our planet.