It’s the conversation that’s coming up in households and conference rooms across the country: are you using AI? Do you like it or are you against it? What about that data center they’re talking about building outside of town? We launched a study to see just how that conversation is evolving and where sentiments are trending about AI today.
AI has become useful enough that many Americans have already made it part of their routines. In our recent research, the majority (81%) said AI helps them get things done faster, while 70% believe it helps produce better work and 59% have a positive view of the technology. Regardless of being for or against AI, it looks more and more like the world will have to learn to coexist with it: nine in 10 believe AI is here to stay.
There are many ways AI solutions and tools can be applied to our personal and professional lives. We asked people to share their own experiences of where AI is helping in their lives and they often described practical ways they were using it.
“AI is making everything easier for us… It helps with simple tasks like summarizing emails or responding without having to write everything.” - Woman, age 25–34
"AI is incredibly useful and improving fast. It boosts productivity, creativity, and learning by processing information faster, explaining complex topics more clearly, and generating ideas from different perspectives." - Woman, age 35–44
What happens when a data center becomes a local issue?
People are less supportive when the discussion turns to the data centers required to support AI. We went to the source of this debate and surveyed U.S. adults living in ZIP codes that already host significant data center developments or are being considered for new projects. While 59% believe data centers are necessary, 46% oppose having one in their own community and 77% oppose development near residential areas.
While half (53%) are concerned about noise pollution from nearby data centers, electricity and water comes up again and again. Seventy percent of people are concerned that increased electricity demand could raise utility bills, while 69% worry about pressure on local water resources. Those aren’t empty concerns either, 72% are willing to make a trade off of slower and less advanced AI-powered digital services if it means reducing pressure on local energy, water and infrastructure.Those concerns come through loud and clear in the open-ended responses. People talk about what’s already happening where they live and how a new data center might affect it. One woman in Texas connected the issue to electricity shortages during extreme weather, rising utility bills and an already limited water supply:
“In Texas, we already face electricity shortages, especially in summer and winter, and data centers consume too much electricity. They are loud, and we end up paying high amounts for electricity and water because of them. We are also running out of water, and data centers exacerbate this issue. Why are our electricity bills increasing?” - Woman, age 45–54
Another person described concerns about local water quality and noise, while others mentioned heat, environmental effects and pressure on the electrical grid.
“I dislike the idea of a data center being built in my community because it would damage our environment and overall health. Specifically, I am concerned about our water becoming undrinkable, our electricity bill increasing exponentially, and constant noise.”
– Woman, age 55-64
These responses help explain why general attitudes toward AI don't necessarily predict how someone will react to a proposed development nearby. AI tools, sites, and features are an abstract digital concept for most… at least until a nearby data center turns that hypothetical into a very real local issue. A low-stakes philosophical debate takes on a completely different tone when it starts to intersect with whatever is already happening in their communities, from electricity costs to water availability.
Interestingly, the intensity of the local debate around data centers doesn’t necessarily mean a high level of awareness. Only about one-third of people living in communities with active data center discussions were aware those conversations were taking place.
Using AI is closely associated with support for data centers
Attitudes toward AI and data centers vary considerably across different groups in our study:
- Major gap between AI users and non-users: As you might expect, experience with AI is indeed associated with much higher support for data centers. Among AI users, 79% said they like AI, 58% believe it will have a positive impact on society and 43% support local data centers. Among non-users, those figures fall to 14%, 6% and 12%, respectively.
- Gen Z is anti-data center: Only 19% of Gen Z adults supported local data centers. Millennials are more open to AI and data centers; ages 25 to 44 reported the highest AI usage and the strongest support for local data centers. AI adoption was 77% among 25-to-34-year-olds and 74% among 35-to-44-year-olds, compared with 40% among Gen Z adults ages 18 to 24.
- Less than a quarter of women support data centers: Seventy-two percent of men had a positive view of AI compared with 46% of women. The gap carried over to data centers, with 42% of men supporting local development compared with 20% of women.
Support for data centers varies considerably depending on who you ask, but experience with AI produced some of the widest gaps in the study. People who already use AI were more than three times as likely as non-users to support local data center development.
Data centers power a lot more than AI, but many don’t realize that
While AI becoming more and more prevalent is a direct reason why so many data centers are being built lately, many Americans aren’t aware that many of the common internet and digital services they use on a daily basis also require data centers. Seventy-one percent knew AI requires data centers, but only 49% made the same connection with cloud file storage. The numbers dropped to 41% for social media, 40% for streaming movies and TV, 31% for streaming music and 29% for messaging apps.
One person, man in the 45-54 age group, who supported data centers points this out in a video response, noting that people tend to forget about all the technology beyond AI that depends on them.
It raises a useful question for future research: How much would people's views change if they had a better understanding of what data centers actually power?
Hearing what sits behind the percentages
We conducted this study as an eight-minute conversational survey on the Rival Technologies platform, combining structured questions with text and video open ends and AI-accelerated Smart Probing. That allowed us to follow up when someone raised a concern and hear more about the experiences shaping their response.
The open-ended responses showed that people can select the same concern for very different reasons. Two people can both select “electricity” as a concern while thinking about very different things: one may be worried about monthly bills while another is thinking about outages or the capacity of the local grid. The same applies to water, noise and other community concerns.
It also gave people room to express some of the uncertainty that doesn't fit easily into an agree/disagree scale. One man who was skeptical about AI summed up a feeling we heard in different forms throughout the research:
“I don’t have a choice on AI being around. It is built into so much already.” - Man, age 45–54
With 90% of Americans already agreeing that AI is here to stay, companies investing in the infrastructure behind it need to understand what people think about development in their own communities. Reach3 can help companies explore local awareness and sentiment, identify the concerns driving support or opposition, test communications and community benefits, and track how views change as projects move forward.
Our conversational research approach combines quantitative measurement with AI-accelerated probing, text and video feedback to capture what people think and give them room to explain why.
Talk to us about building a research program for the communities you need to understand.
You can also Explore the full findings from Reach3 Insights’ AI Infrastructure Paradox study.