Kevin O’Leary, the Shark Tank celebrity venture capitalist also known as “Mr. Wonderful,” lit up social media this month when he suggested that China could be fueling the rapid rise in data center opposition.
O’Leary and his business partners plan to build one of the world’s biggest sites for AI data centers in rural Utah, one that would sit on 40,000 acres and consume more power than Manhattan. Seemingly overnight, furious protests erupted. O’Leary doesn’t believe that this reaction was a spontaneous grassroots movement. He claims that he has found evidence “traced to a network of Chinese money.”
Finally, a reason to check your email.
Sign up for our free newsletter today.
Set aside for a moment whether local residents have reasonable concerns about such a big industrial project. Consider first that AI is indeed a consequential technology with generational economic and collaterally geopolitical consequences, and that the U.S. is the epicenter of the AI boom. It would be profoundly naïve to think that China, as well as Russia, Iran, and North Korea, aren’t trying to stoke opposition on social media and surreptitiously funding opposition groups. Such behavior has a long history in the real world of geopolitics.
China isn’t the only foreign power with deep pockets eager to throw sand in America’s economic gears. In April, the American Energy Institute documented how European foundations and billionaires donated some $39 million to a dozen U.S. organizations that oppose data centers here.
All this interference merely adds fuel to a political firestorm already ignited by America’s professional grievance industry. “New Gas-Powered Data Centers Could Emit More Greenhouse Gases Than Entire Nations,” exclaimed one article asserting that “just 11 data center campuses around the US have the potential to create more greenhouse gases than the country of Morocco emitted in 2024.” Yet another (ridiculously) declaims that Mr. Wonderful’s Utah data center would create a heat island “equivalent to setting off 23 nuclear bombs per day.” And, maybe the deepest cut: “Is AI the new fracking?”
Professional activists have latched on to a new cause célèbre to re-animate an array of flagging grievances. Gallup’s tracking poll shows that scary stories about climate change, for example, seem to have exhausted the public; the issue rarely breaks into the Top Ten list of worries. Affordability, on the other hand, sits right up at the top.
Media coverage frames the enormous power needs of the emerging AI data centers as the reason for an electricity “affordability” crisis. Data centers are also blamed for other trigger issues: water exhaustion, use of farmland, the AI jobs apocalypse, and the depredations of social media.
According to Heatmap News’s project tracking data center opposition, the first quarter of 2026 saw local resistance to almost 100 projects and a record-setting deep-sixing of some 20 sites that represented over $40 billion in potential investment.
In 2017, back when Big Tech was the toast of the town in Washington, D.C., a Gallup poll found that 79 percent of Americans agreed that AI had had “a very or mostly positive impact on their lives.” Cut to today, wherein many people have since learned that both the internet and AI require building data centers just as pharmaceuticals, for example, require building factories. Thus a new Gallup poll finds that 71 percent of Americans “oppose constructing data centers for artificial intelligence in their local area. Talk about zeitgeist whiplash.
It wasn’t as if AI tools weren’t in use nearly a decade ago. AI was already embedded in many consumer services and, at least in the technical community, well-studied and hotly pursued. Gallup noted in its 2017 poll that 85 percent of Americans then used “at least one of six products with AI elements.” But a lot has happened since then not only to make AI more consequential but also far more visible, thus animating the resistance movement.
Consider that Nvidia’s market cap in 2017 stood at about $100 billion and has now blown past $5 trillion. Following the arrival of ChatGPT in November 2022 and the rapid emergence of competitors, the Top Ten publicly traded AI companies alone have a combined $25 trillion market cap. That’s more than double the market value of the entire European stock market. Elon Musk’s net worth was a mere $14 billion back in 2017 and is now rocketing toward $1 trillion.
The reality is that the internet and the cloud, amplified by AI, have become enormously consequential in every sense: physical, economic, and social. The cloud is on track to become the biggest infrastructure humanity has ever constructed and is expanding at a torrid pace. A political reaction to this latest pivot in history was inevitable.
AI will certainly affect the workforce, though claims of mass layoffs already underway are not supported by data. For all the apocalyptic predictions, tracking data from outplacement firm Challenger, Gray & Christmas show no trend of rising layoffs over the past two decades or since the advent of ChatGPT. While it’s true that AI is often identified as one reason for paring employment, of the top dozen reasons, AI has accounted for barely 10 percent of total layoffs year-to-date 2026.
Dislocations are doubtless coming, though. Productivity advances have always been disruptive and aroused political angst. As I reminded readers in my book, The Cloud Revolution, fears of automation replacing jobs have a long history. In the 1960s, Presidents John Kennedy and Lyndon Johnson both took steps to address what JFK called “the major domestic challenge of the Sixties: to maintain full employment at a time when automation, of course, is replacing men.” A Blue Ribbon commission back then even recommended “a guaranteed minimum income for each family.” Sound familiar?
If labor-saving technologies were net job destroyers, though, the unemployment rate should have continually risen throughout modern history. It didn’t. The later collapse in America’s industrial workforce that did take place was driven less by technology than by the insidious and deep expansion of regulatory and environmental rules hostile to heavy and “polluting” industries. These policies drove many companies out of the country, often into the waiting hands of China. Only recently has there been a bipartisan attempt to reverse that.
Meanwhile, the architects of—and investors in—the digital ecosystems are indeed reaping epic profits, or at least market valuations. This dynamic is different from fortunes made because of mandates and subsidies from top-down government diktats, such as for EVs and solar and wind hardware; it’s coming instead from a bottom-up, insatiable appetite that people and businesses have for ever more useful software and AI tools.
As for the outrage over data centers’ excessive or “unfair” use of energy, land, or water? Nearly all of that is demonstrable hype, if not dishonest.
One proposed giga-scale data center site in Indiana would, its opponents hyperventilate, use “millions of gallons” of water annually. But context matters. A corn farm on the same land area would use 200-fold more water. As for occupying otherwise useful rural land: In a few years the nearly 1 million acres estimated to be occupied by sites (not buildings) for data centers will represent an infinitesimal 0.01 percent share of all farmland.
When it comes to visual blight—and we can concede that gargantuan warehouse-scale computers are not things of beauty for most people—where was the spontaneous resistance movement when developers started entirely covering farmland with fields of black solar panels stretching to the horizon?
It’s true that the power appetites of AI-infused data centers are setting new records and creating technical challenges. But the contention that we cannot meet the electric power challenge without punishing consumers is bunk.
Nor do facts support the claim that data centers have caused the recent increase in electricity rates. Nearly all the giga-scale facilities are still under construction and have yet to consume one kilowatt-hour. The prime culprit for rate hikes was the decision by so many utilities and regulators to pursue an “energy transition.”
Of course, like all major construction projects, the build-it phase can be noisy and annoying to local communities. Responsible developers have long sought various ways to ameliorate such concerns, not least reducing local taxes and improving infrastructure. It’s an old model that involves trade-offs. And for the communities that reject the trade, the capital will go, as Walter Wriston famously said, “where it is welcome.”
Still, the resistance is not to be taken lightly.
Foundational revolutions that mark pivots in history are rare, but they are commonly resisted when they occur. It may be of little comfort to those living through the current techno-resistance movement, but history has relevant lessons: many things we now view as commonplace were once radical and disruptive.
For example, vigorous opposition accompanied the emergence of railroads, the automobile, shopping malls, airports, and the chemical industry. While the specific technologies may differ, the suite of criticisms and grievances are similar: sprawl, the hollowing out of legacy institutions or norms, visual blight, pollution, and the inequities of asymmetric enrichment. And throughout history, bad actors often sought to exploit disruptions. Even so, resistance movements always have some reasonable concerns, even if ignored or manipulated in the tumult.
One implicit lesson from earlier resistance movements is that none succeeded in halting the deployment and expansion of the vilified technologies. However, there is one glaring exception, and it is the most recent resistance movement: anti-nuclear activism, which grievously wounded a formerly dominant U.S. industry. For all practical purposes, the U.S. invented the commercial nuclear industry and the technology supply chains that enabled the “taming of the atom.” In the early days, nuclear fission’s astonishing physical, resource, and derivatively environmental benefits were widely celebrated, even if sometimes overhyped with speculations about nuclear-powered commercial ships, airplanes, and even cars. President Richard Nixon, reacting to the energy shock of 1973–74, launched Project Independence, which envisioned 1,000 nuclear power plants by 2000.
A 1964 Gallup poll found that nearly half the country believed nuclear power would benefit mankind. This was just two decades after Hiroshima, and during an eruption of Hollywood horror movies featuring hideous effects of radiation producing gigantic mutant insects and people. (Some might remember Attack of the 50 Foot Woman.) The 1970s anti-nuclear movement eagerly stoked fear of radioactivity and mixed it with the usual “evil big business” tropes. Nuclear resistance metastasized with the accident at Three Mile Island in 1979. The 1986 Chernobyl disaster further fueled the nuclear opposition.
The U.S. commercial nuclear industry succumbed to the media, political, and legal onslaughts. Japan’s 2011 Fukushima disaster served as the final victory for the nuclear opposition, with Europe and Japan rapidly retreating from the peaceful atom.
Though nearly every environmental organization today remains firmly opposed to commercial nuclear energy, overwhelming bipartisan enthusiasm—shared by Big Tech—for resurrecting nuclear energy will likely overwhelm the old anti-nuke movement. But consider the half-century of lost opportunity, and the immense cost of restoring the necessary engineering and industrial nuclear supply chains. In the interim, China vaulted ahead of the United States as the dominant builder of commercial nuclear plants.
Whether the U.S. ends up ceding leadership of artificial intelligence will be, at least in part, determined by how easy or hard it is to build the infrastructure.
Meeting surging demand is requiring massive new spending on power, something totally unexpected by the myopic “energy transitionists,” who would like to use the data center resistance movement as a way to hide the economic consequences of two decades of throttling electric grids and burdening them with costly energy-transition mandates. The transitionists see an opportunity to use Big Tech’s deep pockets to conceal the costs of inadequate power and higher rates—a problem they created in the first place. At least one congressman has said it plainly: “Big Tech should pay for everything.”
If regulators, politicians, and activists try to extract too high of a price from Big Tech to pay its “fair share” of power costs, however, many data centers could simply opt out of public grids and build their own power system on private grids. Many are already doing so. If that becomes the trend, those whose policies have created unaffordable public grids won’t have anyone to hide behind—or any way to lower costs. And that will inspire spontaneous revolt.