AI and Sustainability: Balancing Innovation with Environmental Responsibility

Artificial intelligence is rapidly reshaping how organizations operate—driving efficiency, accelerating insights, and unlocking new forms of creativity. At IA, AI is already enhancing design research, visualization, documentation, and knowledge sharing. These benefits are real and material.

At the same time, AI’s environmental footprint is growing just as quickly. Behind every AI-generated image, meeting summary, and data visualization lies a physical infrastructure that consumes significant electricity, water, and materials, resulting in measurable carbon emissions. As AI adoption expands across industries, responsibly balancing its benefits with its environmental impacts has become a critical sustainability challenge.

The Hidden Environmental Cost of Digital Intelligence

AI systems rely on large-scale data centers that operate continuously to support model training, inference, and storage. In the United States alone, data centers already account for approximately 4% of total electricity demand, with projections suggesting this could rise to 8–12% by 2030 as AI use accelerates. AI-focused data centers can require nearly seven times more electricity than traditional data centers—often exceeding 200 megawatts, enough to power tens of thousands of homes.

Electricity use is only part of the picture. Many AI data centers depend on large volumes of water for cooling, frequently sourced from municipal systems. AI-focused facilities can require 1–5 million gallons of water per day, with a significant portion lost to evaporation or discharged as wastewater requiring treatment. These impacts are especially concerning as new data centers increasingly cluster in regions already facing high water stress.

Communities should be involved, asking questions about the impact of a proposed data center in their venue. Where will it be located? Will it supply its own source of clean energy? Is it going to meet its own power needs or try to force subsidy costs on local energy consumers? What will be the water supply demand, its source, and use?

Carbon Footprint

The carbon implications are both operational and embodied. Operational emissions arise from electricity consumed during AI computation and cooling, while embodied emissions reflect the carbon footprint of IT hardware manufacturing and the shorter lifecycle of AI-focused data center infrastructures—often 10–15 years, compared to 50 years or more for typical commercial buildings.

While energy costs and water demand often dominate the conversation, their impacts on data center footprints extend much further, with growing consequences for both environmental and social health, impacting local communities and long-term quality of life. 

The U.S. has 4,036+ data centers | Source: Data Center Map

Managing Risk While Capturing Value

Responsible AI use is not about limiting innovation—it is about guiding it. AI delivers meaningful business and design value, but that value must be weighed against growing ESG risks, including energy demand, water stress, and Scope 3 emissions growth.

Some data centers are working towards sustainability through multiple techniques, including environmentally responsible construction, renewable energy sources, energy-efficient technologies, and circular-economy practices, but similar to the general building industry, the percentage is very low. For this reason, major tech firms that have set targets for carbon neutrality by 2030 or 2040 are increasingly challenged to meet those goals as energy consumption spirals.

IA’s approach emphasizes:

• Acknowledging AI’s environmental impacts, not obscuring them
• Assessing AI in carbon and water accounting, even as methodologies evolve
• Aligning with providers that demonstrate credible ESG commitments
• Educating employees on both the benefits and tradeoffs of AI tools
• Embedding responsible use principles, governance, and human oversight into AI adoption
• Evaluating mitigation pathways and prioritizing emissions reduction over offsetting

This ensures AI supports IA’s broader commitments under SBTi, CDP, and EcoVadis, and does not undermine long-term climate targets. Our soon-to-be-released 2025 ESG report will highlight our expanded governance for responsible AI use established by AIRIA (IA’s AI Research at IA initiative), along with highlights of our first annual CDP participation, a globally recognized framework for evaluating organizations' climate transparency, risk awareness, and environmental management practices. More detail on CDP and our results will be discussed in our next post.

Designing a Responsible Digital Future

AI will continue to transform design, professional services, and how knowledge flows through organizations. Ignoring its environmental footprint is no longer an option—but neither is rejecting its potential.

Sustainability leadership today requires navigating complexity: embracing tools that improve outcomes while rigorously measuring and managing the impacts they create. By treating AI as both an opportunity and a responsibility, IA is working to ensure innovation aligns with environmental accountability—and that progress does not come at the expense of the planetary systems we all depend on.