Matt Garman, the chief executive of Amazon Web Services, has a straightforward message for Americans who are increasingly uneasy about the spread of data centers: pump the brakes and the country will pay for it. In a lengthy and carefully worded blog post, Garman argues that the wave of local protests against data center construction is not only misguided but, in some cases, deliberately fueled by foreign actors who want to see the United States fall behind in the global race for artificial intelligence. He acknowledges that data centers are easy to dislike. They are enormous, windowless structures packed with servers, running day and night, drawing enormous amounts of power and water, and creating noise, traffic, and construction disruption. But he asks communities to look past the immediate inconvenience and see what these facilities really represent: the physical foundation of the AI era. Without data centers, there is no AI, and without AI, Garman says, the United States loses a historic opportunity to strengthen its economy, protect its national security, and improve the daily lives of ordinary people. He points to breakthroughs in medicine, education, defense, and even entertainment that could become possible with the right infrastructure. With more than one hundred data center moratoriums being debated across the country, he warns that the wrong decision now could put the nation behind for decades, perhaps irreparably. “With any change, there will be important questions raised, but there will also be misinformation and outright lies, and in the age of social media and 24/7 news, myths take hold faster than ever before,” he wrote. For Garman, the choice is not between a data center and an untouched landscape; it is between building the future and letting someone else build it first.
Garman’s reference to misinformation is more than a rhetorical flourish; it reflects a growing anxiety inside the technology industry about how quickly opposition campaigns can spread. In his telling, data center opponents have every right to ask hard questions, but they are also being manipulated by forces that have no interest in honest debate. “In fact, this build out is so important geopolitically that there are widespread reports of various countries intentionally seeding misinformation in the US about data centers to trick us into slowing down,” he claimed. The internet, he argues, has accelerated the spread of myths, making it nearly impossible for facts to catch up with a scary headline. A single viral post about a data center draining a town’s water supply or causing blackouts can do more damage than a hundred community meetings. Garman does not name the countries he believes are behind these campaigns, and his claims are difficult to verify, but they fit into a broader pattern of geopolitical competition over technology. The United States, Europe, and China are all racing to dominate AI, and data centers are the factories of that race. Slowing down their construction is one way to hobble a competitor without firing a shot. By framing the debate this way, Garman is trying to shift the conversation from local grievances to national urgency. He is also trying to give local officials and residents permission to support data centers without feeling that they are betraying their communities. The enemy, he suggests, is not the company building the facility; it is the flood of half-truths and falsehoods that makes reasonable people suspicious. Of course, not all opposition is foreign-inspired, and Garman is careful to say that there are “important questions” to raise. But he clearly believes the pendulum has swung too far toward fear, and that the industry must now fight for hearts and minds with facts, patience, and a willingness to listen.
To make the case more compelling, Garman announced a new initiative intended to show that AWS is ready to be a good neighbor, not just a corporate giant. The program, called Built Together, will allocate one billion dollars over the next five years to communities that host AWS data centers. The exact spending will depend on what each community wants, but Garman says the money is likely to go toward education, job training, energy affordability, and water and energy conservation. It is a significant figure, but also something of an olive branch to local leaders who have grown tired of big tech promises that never seem to materialize. Alongside the financial commitment, Garman revealed that AWS has stopped using nondisclosure agreements when seeking approval for new data centers from government agencies. NDAs have long been a source of friction, with local officials complaining that they were asked to sign away their ability to discuss projects with constituents. Like Microsoft, which made a similar move, AWS is trying to signal a new era of openness. Garman says the company now provides information about its plans as early as possible, listens to concerns, and adjusts designs to fit local conditions. The name “Built Together” is itself a message: data centers should not be imposed on communities, but developed with them. He emphasized that the money is not a bribe and will not be controlled by executives in Seattle; it will be shaped by the people who actually live near the facilities. In an industry that has often treated local opposition as a public relations problem to be managed, the shift is notable. But it also sets a high bar. If AWS wants to be treated as a partner, it will need to act like one over the long term, not just when new projects are on the line.
Perhaps the most emotional battle is over the environment. Data centers are notorious for guzzling electricity and water, and the large backup diesel generators outside these facilities have become a symbol of pollution for activists. Garman pushes back hard. He argues that data centers actually use very little water relative to other industries, and he offers a striking comparison: American golf courses use about two hundred times as much water as all of Amazon’s data centers combined. His point is not to make golfers the enemy; it is to show that data centers are being singled out when far thirstier industries have never faced moratoriums. On energy, he says data centers are not the primary cause of rising electricity prices, and he points to efficiency improvements that have helped the industry do more with less. On diesel generators, he dismisses the idea that they are a major pollution source, noting that they are designed for emergency use, not routine operation. He also reminds readers that AWS has invested heavily in renewable energy and is committed to water conservation projects around the world. Still, these arguments are unlikely to silence critics who see data centers as industrial infrastructure in residential areas. A golf course, after all, is mostly green space; a data center is a box of metal and concrete. But Garman is asking communities to weigh the actual data against the imagined horror, and to recognize that the environmental picture is more complicated than a meme or a viral tweet suggests. He is not denying that data centers consume resources. He is saying that every major infrastructure project has costs, and that data centers should be judged fairly, in context, rather than held to an impossible standard that no industry could meet.
Garman also makes a forceful economic case, and it may be the most persuasive part of his post. Data centers, he says, bring considerable tax revenue, positive economic impact, and a large number of good jobs to the communities where they are built. He is not only talking about the engineers and technicians who maintain the servers; he is also talking about the thousands of construction workers who pour concrete, bend steel, and install electrical systems. These are not abstract corporate jobs. They are the kind of employment that can transform a small town, supporting families and local businesses like diners, hotels, and hardware stores. “These jobs benefit families in these communities and are a boon for local small businesses,” Garman wrote. He says the average wages paid by data center projects far exceed local and national medians, and the number of jobs is often very significant relative to the size of the surrounding community. For a county that has lost a factory or a coal mine, a data center can look like a lifeline. That is why so many local leaders support data centers even when their constituents do not. But the economic argument is not as simple as Garman makes it sound. Recent research in the United Kingdom, for example, found that only about a quarter of the jobs promised in upcoming data center projects will actually materialize. The environmental think tank Verdant analyzed projected roles and concluded that data centers would create around 10,400 workers, far fewer than the roughly 40,000 that industry groups like TechUK have touted. The discrepancy matters because job creation is one of the most powerful arguments in favor of data centers. If that promise is overstated, the entire case becomes weaker. Garman insists AWS is trying to give a realistic picture, but he also points out that data centers are not meant to be employment centers in the same way as a factory; they are highly automated facilities that create value in other ways. The tax revenue alone can help pay for schools, roads, and public safety, and that is something communities feel immediately. The challenge for AWS is to convince skeptical residents that these benefits will actually reach them, rather than being captured by shareholders and distant corporations.
Despite Garman’s best efforts, the numbers suggest that his message is not getting through. A poll conducted last month by the University of Massachusetts Amherst found that only 11 percent of Americans would support the construction of an AI data center in their local community, while 65 percent would oppose it. The resistance cuts across party lines, which is rare in a deeply polarized country. Raymond La Raja, a political science professor at UMass Amherst and co-director of the poll, put it bluntly: “The remarkable thing is that we can’t find a constituency that actually wants these data centers. The politics of AI may be somewhat polarized, but resistance to putting its infrastructure next door crosses the partisan divide.” That is a striking rebuke to an industry that has spent years trying to convince the public that data centers are clean, quiet, and beneficial. The backlash extends far beyond the United States, with data centers facing moratoriums, lawsuits, and community revolts in several countries. Garman’s blog post is unlikely to change that overnight. It may, however, have a more modest effect: it could give local officials and prospective host communities a reason to pause before saying no. It could also put pressure on other tech companies to follow AWS’s lead in dropping NDAs and offering tangible benefits. But the deeper issue remains a crisis of trust. People have heard too many promises from big tech, and they have seen too many examples of infrastructure being built without adequate oversight. Garman can argue that data centers are the future, that foreign adversaries are spreading lies, and that the benefits outweigh the costs. None of that matters if the person standing outside a proposed site believes that their property value will drop, their water bill will rise, or their children’s air will be dirtier. In the end, the data center debate is not really about algorithms, energy grids, or national security. It is about fairness, control, and the feeling that ordinary people have a say in the places they call home. Garman has made his case, and he has offered a billion dollars to back it up. Whether that is enough to bridge the gap between the industry’s vision and the public’s fear is still very much an open question.