NTT DATA Urges Action for Sustainable AI in New White Paper “Sustainable AI for a Greener Tomorrow”
NTT DATA, a global leader in digital business, AI, and technology services, has released a new white paper underscoring the urgent need to integrate sustainability into every stage of artificial intelligence (AI) development and deployment. The report, titled “Sustainable AI for a Greener Tomorrow,” calls on organizations worldwide to view sustainable AI not just as an ethical responsibility but as a strategic opportunity to create long-term value, strengthen resilience, and conserve critical natural resources.
As AI adoption accelerates across industries, so too does its environmental footprint. Training large language models (LLMs), maintaining inference pipelines, and supporting continuous AI operations demand enormous amounts of electricity and computing infrastructure. According to NTT DATA, researchers now estimate that AI workloads could account for more than half of all data center power consumption by 2028—a figure that highlights the scale of the sustainability challenge.
Beyond energy, AI systems also contribute to rising water consumption, as data centers require significant volumes of water for cooling. Additional environmental pressures stem from electronic waste and rare earth mineral extraction used in the production of chips and hardware components.
Despite these concerns, NTT DATA emphasizes that AI itself can play a vital role in mitigating the problems it helps create. “The resource consequences of AI’s rapid growth and adoption are daunting,” said David Costa, Head of Sustainability Innovation Headquarters at NTT DATA. “But the technology also empowers groundbreaking solutions to many of these same challenges. AI can optimize energy grids, reduce emissions, model environmental risks, and enhance water conservation. The key is to embed sustainability from the outset—making it a fundamental design principle, not an afterthought.”
Key Insights from the White Paper
The NTT DATA report presents a structured framework for advancing sustainable AI through measurable and collaborative action. Four central themes emerge as priorities for both organizations and policymakers:
1. Expanding from Performance to Green Priorities
The company advocates shifting the focus from traditional AI metrics—like speed and accuracy—to broader sustainability indicators. Efficiency should be considered a design cornerstone rather than a trade-off. By integrating green priorities into model architecture, software design, and deployment strategies, organizations can drive innovation while reducing carbon intensity.
2. Quantifying Environmental Impact
To make sustainability tangible, NTT DATA stresses the need for standardized, verifiable metrics to track AI’s energy use, carbon emissions, and water footprint. Emerging benchmarks such as the AI Energy Score and Software Carbon Intensity (SCI) for AI are promising tools to incorporate into governance, procurement, and compliance frameworks. Quantification enables accountability and continuous improvement across the AI ecosystem.
3. Adopting a Lifecycle-Centric Approach
A truly sustainable AI strategy demands a holistic view—one that considers environmental impacts from the extraction of raw materials to system deployment and eventual disposal. This lifecycle thinking encourages longer hardware lifespans, optimized cooling systems, and the application of circular economy principles. Steps such as modular system design and component reuse can drastically reduce e-waste and emissions.
4. Promoting Shared Accountability
Sustainability cannot rest on a single actor. Hardware manufacturers, cloud providers, developers, policymakers, investors, and consumers must all share responsibility. NTT DATA calls for cross-sector collaboration to achieve systemic change, ensuring sustainability goals align across supply chains and regulatory environments.
Barriers to Progress—and Best Practices for Change
While awareness of AI’s environmental toll is growing, progress remains uneven. Many organizations still conduct fragmented sustainability assessments, focusing on a single dimension such as energy or emissions, while overlooking other critical factors like water use or material scarcity. Moreover, even when sustainability targets exist, few firms have practical strategies for implementing them across the entire AI lifecycle.
To bridge this gap, the NTT DATA paper proposes a series of best practices to operationalize sustainable AI:
- Implement green software engineering techniques to minimize energy use and resource consumption during AI development and operation.
- Schedule AI workloads during periods of peak renewable energy availability or in data centers located near renewable sources.
- Leverage remote GPU services and on-premises AI solutions to optimize infrastructure efficiency.
- Prioritize modular and upgradable components to reduce e-waste, and extend hardware life through refurbishment, reuse, and responsible recycling.
According to NTT DATA, these practices represent a foundation for systemic change. The company argues that a comprehensive, end-to-end redesign of the AI lifecycle is not only possible but essential to ensure that AI’s positive potential does not come at the expense of environmental integrity.
NTT DATA’s call to action is clear: the journey toward sustainable AI demands both innovation and accountability. By integrating sustainability goals into every level of AI governance—from model training to supply chain management—organizations can ensure that the digital future remains both technologically and ecologically sound.
“AI can either accelerate the global sustainability crisis or help solve it,” Costa added. “The difference lies in how responsibly we innovate.”
NTT DATA invites enterprises, researchers, and policymakers to collaborate in advancing sustainable AI practices and to explore the insights presented in the white paper.
Source Link : https://www.businesswire.com/