Amazon Web Services (AWS) said on Thursday, June 11, that its data centers around the world withdrew about 2.5 billion gallons of water last year to cool the servers that power its cloud-computing and artificial-intelligence businesses. The figure was detailed by AWS executives including Kerry Person, vice president of data center operations, and Will Hewes, the company’s water stewardship lead. It is one of the clearest pictures Amazon has ever provided of the water footprint behind the global computing boom.

The disclosure matters because Amazon has long faced criticism for providing limited information about data-center water consumption. Rivals Microsoft and Google have published water-use figures for years. Amazon had largely focused on efficiency metrics rather than total withdrawals, and earlier this year investors filed resolutions urging major technology companies to provide greater transparency. Thursday’s announcement is Amazon’s most direct response yet.

Amazon is presenting the figure as evidence that its operations are highly efficient. The company says it uses approximately 0.12 liters of water per kilowatt-hour of computing, compared with an estimated industry average of 0.84 liters per kilowatt-hour. According to AWS, that makes its operations roughly seven times more water efficient than the average data-center operator. The company said outside auditors reviewed the calculations and that water withdrawals at facilities Amazon directly owns and operates declined about 2% year-over-year, even as its global footprint expanded.

The company attributes much of the reduction to its cooling strategy. Data centers generate enormous amounts of heat, and many operators rely heavily on evaporative cooling systems that consume significant amounts of water. AWS says its facilities use outside-air cooling about 90% of the time, relying on fans to move air through server halls. Water cooling is generally used only when outdoor temperatures exceed roughly 85 degrees Fahrenheit. The company also adjusted operating temperatures within its facilities to further reduce cooling demand.

The disclosure arrives at a sensitive moment for the industry. In Amazon’s home region, the Seattle City Council this week unanimously approved a one-year emergency pause on new large data-center developments within the city. The action reflects growing concern among local governments over the water, electricity, and land demands created by artificial-intelligence infrastructure.

Person said community reactions are often different from what critics expect.

“As we’ve been engaging with our local communities, they’ve been very pleasantly surprised about how little water we are using,” he told reporters.

Not everyone agrees. Simon Hans Edasi, a Seattle-area data scientist who studies data-center development and water resources, has raised concerns about Amazon’s planned $4.8 billion campus in Burbank, Washington, near the Columbia River. He argues that the industry is increasingly expanding into eastern Washington and other regions where water supplies are already under pressure.

Several recent studies have found that a significant share of new U.S. data-center construction is occurring in areas experiencing varying degrees of water stress. Those concerns have fueled permitting battles, project delays, and in some cases the cancellation or relocation of major developments.

For companies investing tens of billions of dollars in AI infrastructure, community opposition is becoming a material business risk. Delays in permits and approvals can significantly increase costs and slow expansion plans.

Amazon says its long-term answer is its Water Positive by 2030 commitment, first announced in 2022. The company says it has completed approximately 75% of the work needed to achieve that goal and currently replenishes about three gallons for every four gallons it uses.

According to Hewes, the strategy focuses on three priorities: reducing water consumption, replacing drinking water with treated wastewater whenever possible, and investing in local replenishment projects. Those efforts include repairing leaking municipal infrastructure, restoring watersheds, and supporting agricultural irrigation programs that use recycled water.

Microsoft has announced similar goals, including a pledge to improve water efficiency by 40% by 2030 and replenish more water than it consumes in the regions where it operates.

As artificial intelligence drives unprecedented demand for computing power, technology companies are increasingly competing not only on performance and scale, but also on environmental impact.

Amazon also highlighted a broader industry statistic, noting that global data centers account for approximately 0.5% of industrial water use worldwide. Whether that argument satisfies communities increasingly wary of large-scale AI development may ultimately be decided one project at a time.

JBizNews Desk — Technology

© JBizNews.com All Rights Reserved. Reproduction or distribution without written permission is prohibited.