By Adam McMillen

Artificial intelligence has been a game changer in engineering. This includes IMEG, which was an early adopter of AI tools to assist our engineers in delivering sustainable design for long-term, positive environmental outcomes. As a Top 20 Green Design Firm in the U.S., however, we understand using AI comes with a carbon footprint all its own.

AI relies on energy-intensive high-computing data centers that consume vast amounts of electricity to run the computations and large amounts of water to cool the servers. Therefore, every prompt or chat message has a carbon footprint. With 3,000 employees at IMEG, modest individual daily AI usage across the firm adds up to tens of thousands of gallons of water, numerous megawatt-hours of electricity, and several tons of carbon emissions annually.

So the question for us became: Can we deliver highly impactful, sustainable designs resulting in carbon reductions that far outweigh the AI carbon footprint of the design process? The answer is a definitive yes.

We track our overall AI carbon footprint and have found that leveraging AI in our design work has the potential to save more than 10,000 times the carbon it consumes. We achieve this in two ways. First, we minimize our carbon footprint by having our proprietary AI tool utilize OpenAI models hosted by Microsoft—which has pledged to source 100% renewable energy for its data centers by 2030. Compared to the U.S. grid average, this effectively halves the emissions produced by each AI interaction.

The other and more impactful strategy is due to the speed with which AI analyzes our own project data and assists our engineers with repetitive tasks. This enables us to increase our carbon analysis of multiple design options for new buildings from about 100 projects annually to more than 1,000 projects annually—quickly providing a substantially larger number of clients with the design that will best reduce their carbon footprint.

Based on industry carbon data, if AI could be leveraged to guide just 5% reductions across 1,000  projects, IMEG could help these clients collectively avoid about 13,000 tons of CO₂ per year in operations and 40,000 to 55,000 tons per year in embodied carbon at construction. This far outweighs the few tons of carbon emissions annually created by IMEG’s use of AI.

While every AEC firm’s AI strategy will differ to some degree, the key to achieving the greatest real value is to ensure they are creating the least amount of carbon emissions possible during design and construction and saving much more in the final product. In this way, engineers, architects, and contractors combined can exponentially reduce the carbon footprint of the entire built environment.