In a recent report on AI Maturity, the global IT consultancy Accenture found only 12% of global organizations polled are advanced enough with AI to qualify as ‘high achievers.’  One of the few non-financial characteristics distinguishing that sliver of companies from the majority, was a commitment to their customers in the form of driving sustainability with AI.

Despite the added complexity and investment this would require, these organizations are embracing an early-adopter’s mission to holistically serve the best interest of their customers. AI is not likely to be a panacea for sustainability any more than it is for other business problems. However, once its underlying costs are accounted for—which include its own environmental footprint—it is expected to provide extremely valuable incremental benefits.

Learning to Manage AI’s Environmental Impact

In a 2020 Forbes article, Glenn Gow explored the environmental costs of running sophisticated AI computing. He cited a study by the University of Massachusetts that included an analysis of the costs of training AI models to perform natural language processing (NLP). UMass researchers calculated that the cost of running and cooling servers for this type of activity produced 5 times more carbon output than the lifetime emissions of an American car, 57 times more than a human in a year, or 300 round-trip flights between San Francisco and New York City. This volume of output is staggering, but also likely to be a fraction of what we will witness as more businesses turn to AI-based solutions in a ChatGPT world.

Organizations that have already made sustainability gains leveraging AI understand they must breach the threshold costs of running the technology before the value of using it to discover cost savings elsewhere can be realized.  

Large Enterprises Must Lead the Way

According to a January 2023 report by McKinsey, the use of renewable energy is a critical component for many tech companies when deciding which data centers they will do business with, or when designing their own centers. A single hyperscaler’s data center can use as much energy as 80,000 U.S. households each year, according to the report. At the same time, the U.S. currently accounts for 40% of the global market for data centers, which implicates the entire domestic data center economy as a sizable polluter, and one that is expected to remain a locus of concern over energy efficiency in the future.  

Major cloud service providers (e.g., AWS, Microsoft Azure, Google Cloud, etc.) are already intimately familiar with the landed costs of running compute-intensive data centers and the pressure to respond to their customers’ environmental concerns while keeping their costs as low as possible. And they have been working steadily to fold sustainable practices into their operations, in part by negotiating renewable resources to power their centers. Still, they face numerous challenges, and adding AI to their clients’ computing load will create even more.

Research suggests that some combination of large-scale size and tech-forward companies will be best equipped to lead the way with AI for sustainability missions, but it’s unclear where that leaves the rest of the business sector.

Ensure Positive Impact

As a 2021 Boston Consulting Group (BCG) report found, large companies are ideally situated to leverage the sustainability benefits of AI because of their large data sets and the scale at which they would then implement solutions. On average, BCG found that AI could help their clients achieve a 5-to-10 percent reduction in carbon emissions. That may only be an incremental step towards the goal of a 50% emissions reduction set forth in the 2016 Paris Agreement, but it is a step worth taking.  Notably, the report did not mention the carbon footprint of the AI itself included in their calculations.

If the largest and most technologically advanced organizations are, indeed, the best positioned to apply AI towards sustainability, then well-intentioned smaller businesses would be wise to thoroughly research their options and look to partners that can match their commitment, scale, and sustainability focus.