With the increasingly popular use of AI in the enterprise, it becomes crucial to ensure that these technologies are harnessed in a climate-neutral way.

Lakshmanan Chidambaram, President, Americas Strategic Verticals

March 15, 2024

4 Min Read
green earth with windmills churning
ZoonarGmbH via Alamy Stock

In an era dominated by the exponential rise of artificial intelligence evolving from predictive to generative and used increasingly ubiquitously from autonomous vehicles to virtual assistants, AI is becoming a popular innovation engine. Underscoring the prevalence and promise of the technology, according to PwC, AI has the potential to deliver total economic activity of approximately $15.7 trillion by 2030 globally

Yet, hidden beneath the sheen of progress lies a growing concern: The substantial carbon footprint AI technologies leave behind. As generative AI gains more popularity, it becomes crucial to ensure that these technologies are harnessed in a climate-neutral way. This requires investment in energy innovation, in addition to putting efforts towards AI development. 

An enormous amount of electricity is needed to train AI algorithms, as well as build and operate the hardware on which these algorithms run. In 2019, MIT researchers found that training an AI model could emit more than 626,000 pounds of CO2, which was nearly five times the lifetime in emissions of an average American car. A Cornell University study estimates global AI demand may account for 4.2 - 6.6 billion cubic meters of water withdrawal by 2027, much of which evaporates, rendering it unusable. This trajectory jeopardizes climate goals, necessitating the recalibration of strategies to combat climate change while bridging technology and sustainability for mutual progress. 

Related:A Portal into Sustainability

As AI systems become more prevalent, they have the potential to both exacerbate environmental challenges and offer solutions. It is imperative to strike a balance between harnessing AI’s power for innovation and mitigating its negative sustainability effects. A recent Boston Consulting Group report shows that AI has the potential to mitigate 5-10% of global greenhouse gas emissions by 2030. 

For instance, if an AI system prioritizes economic growth over environmental protection due to its susceptibility to biased decisions from flawed or incomplete data, it could prioritize short-term financial gains, risking environmental sustainability. Let’s consider an AI-powered supply chain management system within a business. If the AI algorithms solely prioritize cost reduction and speed of delivery without factoring in environmental impact, consistent recommendations of suppliers or transportation routes that offer the lowest costs but have higher carbon footprints may be expected. This approach might neglect environmentally friendly suppliers or shipping methods, exacerbating the carbon footprint of the business. 

Related:Special Report: What's Next for the GenAI Market in 2024?

Transparency and Accountability 

Addressing the challenge of AI’s impact on sustainability in businesses involves multiple strategies. Ensuring high-quality, diverse data for AI training minimizes biases. Transparency in algorithms aids in identifying and rectifying biases, much like a supply chain AI system prioritizing both cost efficiency and low environmental impact. 

For instance, businesses can develop a robust sustainability framework where both carbon footprint and performance metrics (cost, lead time, quality, service level) are weighted to maintain efficiency and responsiveness across the entire supply chain or operational process. Additionally, educating stakeholders about environmental considerations in AI fosters conscious decision-making. Regulatory standards mandating eco-conscious AI use further solidify these efforts and should be embraced. 

Use of Renewable Energy 

The adoption of renewable energy sources, such as solar, wind, and hydropower, can significantly reduce the carbon footprint of AI operations. Renewable energy curtails greenhouse gas emissions and offers a more sustainable and scalable power solution for the data centers and hardware infrastructure that support AI systems. Moreover, the US offers various policy incentives to support the expansion of green energy infrastructure to promote renewable energy resources and reduce greenhouse gas emissions. 

Related:How Climate Change is Changing Cyber Resilience Plans

Various federal and state agencies offer grants and funding opportunities for renewable energy research, development, and deployment. The federal tax credits also encourage the development of renewable energy projects, with the investment tax credits primarily supporting solar energy. At the same time, the production tax credit applies to wind, geothermal, and other renewable resources. 

Green AI Algorithms 

Green AI algorithms aim to reduce AI’s carbon footprint by tackling the problem at the data level, encompassing various strategies to reduce computational resources in AI management. 

One example involves model optimization techniques such as knowledge distillation. This process involves transferring knowledge from a large, resource-intensive AI model to a smaller, more lightweight one, reducing the computational burden while maintaining performance. Similarly, pruning techniques identify and eliminate redundant connections within neural networks, cutting down on computational requirements without compromising accuracy. 

Another avenue involves energy-aware scheduling, which strategically allocates tasks to optimize energy usage in AI-modeled data centers. 

A Call for Solutions 

To ensure AI and environmental sustainability coexist harmoniously, organizations must embed sustainability into the core values and long-term strategies. This means setting clear environmental goals, measuring impact, and reducing carbon footprints. 

To fully realize the vast benefits associated with the expanded use of AI, enterprise-level approaches as strategies must be viewed through the lens of sustainability. Sustainable AI must not be relegated to the “wishful thinking” category but must be pursued as a corporate imperative. Embracing eco-friendly algorithms, renewable energy, and responsible practices, AI can thrive while preserving our planet. 

Let’s code a future where technological progress doesn't come at the expense of the earth. Through conscious choices and collaborative efforts, we can shape an AI-powered world that sustains and rejuvenates our planet. 

About the Author(s)

Lakshmanan Chidambaram

President, Americas Strategic Verticals, Tech Mahindra

Lakshmanan Chidambaram (fondly known as CTL) is president – Americas strategic verticals and has been the face of Tech Mahindra's largest business unit since 2016, overseeing strategy, innovation and expansion of the organization's multibillion-dollar footprint across industries in the region. He is responsible for a diverse range of industries, including banking and financial services, insurance, manufacturing, retail and consumer packaged goods, travel and transportation, healthcare and life sciences, energy and utilities, and the public sector, consistently delivering industry-leading growth and profitability. 

Never Miss a Beat: Get a snapshot of the issues affecting the IT industry straight to your inbox.

You May Also Like

More Insights