Artificial Intelligence for Sustainability

Sustainability at the Speed of AI:
Unlocking Responsible Transformation

The explosion in Gen AI may be sustainability’s blessing and curse. Here’s how companies can use AI to scale climate initiatives while accounting for its impact.

Sponsored Content | This article is sponsored by Capgemini | By Franco Amalfi and Christopher Scheefer | Published 3 months ago | About a 6 minute read

Image: Frolopiaton Palm

Organizations today must navigate the dual pressures of digital acceleration and climate responsibility to stay competitive.

But this presents a dilemma: Generative artificial intelligence, one of the emerging technologies with the greatest potential to drive digital transformation, has a tremendous impact on the environment.

The perception that Gen AI can produce a cavalcade of business benefits expanded its use rapidly. But this techno-optimism is tempered by the realization that the current trajectory could exacerbate the climate crisis and further strain our natural resources.

Gen AI, which can produce new content based on patterns in existing data, requires vast energy consumption and water usage. This technology cannot function without the production of graphics-processing units — fueling both the mining of rare earth metals and greenhouse gas (GHG) emissions.

Capgemini believes that embracing Gen AI and protecting the environment aren’t mutually exclusive. However, we acknowledge this quandary and aspire to face it head on — and we encourage others to do the same.

Deploying Gen AI-powered solutions in ways that optimize an organization’s energy use while mitigating its own impact will require careful planning and the adoption of ethical frameworks. Approached intentionally, the integration of Gen AI can supercharge eco-friendly business practices — something we call “sustainability at the speed of AI.”

Sustainable AI: Problematic paradox or promising proposition?

According to an analysis by software engineering website Baeldung, training ChatGPT-5 is expected to consume 3,500 megawatt-hours — enough energy to power roughly 320 average US homes a year. Simultaneously, the global AI market is projected to hit $4.8 trillion by 2033 — up from $189 billion in 2023.

The relentless demand to deploy LLMs at scale poses a significant problem for organizations with sustainability commitments.

The Capgemini Research Institute report, Developing Sustainable Gen AI, examined the technology’s environmental impact and the need to develop sustainable usage practices. The team conducted a survey of 2,000 senior executives at organizations with annual revenues exceeding $1 billion that already have Gen AI initiatives underway.

Only 12 percent of executives surveyed said their organizations measure Gen AI’s environmental footprint, and only 20 percent rank the footprint as a top-five factor when selecting or building Gen AI models. Nevertheless, 48 percent say their Gen AI use has increased GHG emissions.

This underscores the urgent need for more responsible approaches to AI development. Companies and countries alike have falled short on many climate commitments, which often rank second to financial priorities. Should we see AI as the opportunity to finally fulfill our goals or another challenge to compel even more backpedaling?

The good: Accelerating sustainability with AI

Developing environmentally responsible AI for business requires a two-pronged approach: AI should improve the sustainability of operations, and AI itself should be handled sustainably. Here’s what that looks like.

Organizations can use Gen AI tools to embed sustainability throughout countless functions — for example, scenario modeling for environmentally informed decision-making, ESG reporting for compliance and transparency, process optimization for efficient energy use, supplier analysis for Scope 3 emissions reduction, eco-design assistance for cleaner product development, and so forth.

Many Capgemini clients using AI and data for sustainability purposes have reported greater employee retention, emissions reduction, supply chain acceleration, water conservation and cost savings.

Agentic AI, which makes decisions and acts autonomously, can scale these new capabilities with precision and speed. Embedded into daily operations, AI agents can monitor pollution and optimize procurement while humans provide ethical judgment, strategic direction and creativity.

More and more studies forecast that AI will make or augment half of all business decisions by 2027 — but human oversight will remain crucial.

The bad: Grappling with AI’s environmental impact

But the predicament lies in AI’s toll on our natural resources.

A lack of transparency from major LLM providers regarding the impacts of AI use makes it difficult for organizations to incorporate this information into their net-zero strategies.

Cornell University recently published a peer-reviewed paper by our Capgemini colleagues, who proposed a comprehensive methodology for estimating the environmental impact of a company’s AI portfolio. Their methodology breaks down the impacts of AI solutions at the company level into interconnected models, so initiatives can be aligned with sustainability goals.

“To minimize environmental risks, all actors of the AI value chain — including hardware manufacturers — must actively contribute to responsible deployment and usage,” the authors wrote. “Achieving success requires greater transparency through information sharing among stakeholders, including environmental impact data and optimization methods. Without coordinated effort from model providers to end users, environmental impacts will significantly increase.”

To operationalize these strategies, Capgemini developed a framework for integrating sustainability into every phase of AI development and deployment. This includes diagnostics to assess environmental impact, design principles for sustainable AI, training for data science teams, and governance models to ensure ethical and transparent use.

This approach — which supports fast prototyping, pilot implementation, and enterprise-wide scaling — is designed to help organizations move from isolated use cases to holistic transformation strategies. These strategies contribute to the larger goal of lowering the impact of enterprise information technology, which contributes greatly to the world’s carbon footprint.

Real-world benefits across industries

Capgemini’s responsible AI framework is already helping clients identify where AI tools can drive greater efficiency and sustainability:

A biopharmaceutical company identified more than 80 use cases for enhancing digital operations with AI.
A pet food company used Gen AI to reduce product development time by half.
A mining company used drones and a geographic information system — both powered by AI — to automate land monitoring, saving hundreds of thousands of euros annually.
In healthcare, a business improved patient support with AI assistants — reducing response times by more than 80 percent.
And an aerospace company used AI to optimize manufacturing and forecasting, enhancing competitiveness and sustainability.

These examples demonstrate that AI can drive both environmental and financial performance when applied responsibly. But monitoring the environmental impact of the AI-powered solutions in each instance remains critical.

Benefiting both business and planet

By embedding sustainability into AI and AI into sustainability, enterprises can unlock new value, reduce their environmental footprints, and lead in environmental stewardship.

A strategic business transformation partner can help companies develop a structured approach for identifying opportunities, prioritizing initiatives, developing business cases, and scaling solutions that — along with tools for diagnostics, design, implementation and governance — can help organizations move quickly from concept to value realization.

The convergence of AI and sustainability is reshaping the business landscape in a way that’s teeming with challenges and opportunities.

Learn more about Capgemini’s vision for developing sustainable Gen AI practices, and don’t hesitate to reach out to keep exploring solutions.

Franco Amalfi

Director, Sustainability Strategic Initiatives and Partners, North America
Capgemini

Christopher Scheefer

VP, Global Data & AI Lead for Energy, Utilities and Sustainability
Capgemini

Christopher Scheefer is the Global Data & AI Lead for Sustainability at Capgemini, where he helps enterprises harness data, generative AI and automation to accelerate their sustainability transformation. With decades of experience across energy, utilities and manufacturing, he specializes in designing purpose-led, AI-powered solutions that reduce carbon impact, improve resilience and unlock new business value.

Tags:

Capgemini