“There’s so many articles that are just misinformed. And it’s just not true,” Brandon Oyer, head of energy and water for the Americas at AWS, said on one panel. “There’s a very common misconception that there’s many, many hours of the year, if not all hours of the year, we’re using water, and that’s just not true,” he later added.

Execs pointed to AWS’s plans to reduce water via engineering optimization, local replenishments and intervention projects. Per Amazon’s Oyer, a data center in Northern Virginia uses a million gallons of water 4% of the year. “A million gallons, it sounds like a lot of water. But to put it in context, a golf course is going to use about seven to eight times that,” he said.

Back in 2014, data centers used up to 21.2 billion liters [5.6 billion gallons] of water, with most being internal data centers at 64%, according to an Energy Department report. In 2023, total water use increased to 66 billion liters [17.4 billion gallons]. Now, hyperscale and colocation centers (or owned and shared infrastructure) are the big water users at 84%. Internal data centers account for 12% thanks to better water-saving methods.

Other panel topics from the Amazon event focused on AI’s impact on the workforce and the ethical use of AI, in addition to the efforts it claimed it’s making to create community programs. Amazon laid out workforce development programs, affordability and access services and other internal investments for AI usage.

Bloomberg last November reported Amazon data centers topped 900 as the AI frenzy continues. (Amazon did not respond to a follow up request for comment in time for publication.)

Impacts to climate change and the human workforce have been sticking points for many, who are critical of AI platforms and companies using AI tools. Brands and platforms alike have been retooling their marketing messaging to re-center human agency as a result.

“When we see brands use AI as a shortcut or as a copout, that’s where I see the descent and the negative sentiment comes through,” said Graham Douglas, co-founder and creative director at Gus, a New York City-based creative strategy company. He added, “The thing that works is finding the humanity in the technology.”

What’s notable is the timing. Over the past year, Amazon has been the center focus of several news reports regarding its data centers, climate footprint and workforce. 

Last fall, AWS struck a deal with OpenAI, which would allow OpenAI to scale its infrastructure. CNBC reported the deal meant using existing AWS data centers and eventually building out additional infrastructure for OpenAI. Around the same time, Reuters reported Amazon’s plans to cut 30,000 corporate jobs to reduce bureaucracy and increase the use of AI. Then, more than 1,000 Amazon employees signed an open letter to express concern about the company’s AI development.

Despite the increase in consumers’ use of AI — at least 60% of people now use AI weekly, per Klaviyo’s recent research report — scrutiny around AI has intensified due to its environmental effects and perception across social media, privacy and public policy. As companies clamor to roll out AI-powered tools, they’re also weighing public perception of doing so.

“We’re in very much the early stages of things. We have choices and decisions to make today in terms of the way in which we invest and how we see what may come, but also start to prepare,” Diya Wynn, principal responsible AI lead at AWS, said during an editorial exchange panel. Wynn spoke on how the company and consumers are thinking about the future of AI use.

Amazon isn’t alone in facing public skepticism. OpenAI agreed to a deal with the Pentagon last month, giving the Pentagon access to its AI models. Rival Anthropic rejected the deal and OpenAI now faces criticisms over its partnership from staffers and customers. Even brands like Coca Cola, McDonald’s and Svedka vodka have faced public backlash.

It’s unlikely the AI hype bubble will burst anytime soon as companies latch onto promises of efficiency and scale. In the meantime, companies are turning toward messaging around responsible, human-first use to shape public perception.

“You’ve got to find the culturally relevant human reason for being to really resonate,” Douglas said.