In a nation on the precipice of a technological revolution, there are few states quite like Virginia, where hundreds of warehouses full of powerful and power-hungry computers have earned one region the nickname “Data Center Alley.”
So it was all the more surprising when Virginia lawmakers decided in late June to take aim at the industry that they had long helped to incubate. Capping a year of hard-fought negotiations, the state became the first in the country to impose a small tax on data centers based on the vast amounts of electricity they consume.
For the trillion-dollar behemoths that own or lease data centers, including Amazon and Microsoft, the tax represented a fraction of what they might someday earn from artificial intelligence. But for Virginia it is expected to be a windfall, supplying up to $600 million in revenue in the coming year, according to L. Louise Lucas, a Democratic state senator. The rapid proliferation of data centers had raised new economic and environmental costs, she said, and the tech industry needed to shoulder more of that burden.
“How do we let the wealthiest companies in human history come to the commonwealth,” Ms. Lucas said, “and not play by the same rules as everyone else?”
The debate in Virginia is part of a larger reckoning now underway in American politics. From the smallest city councils to the corridors of power in Washington, policymakers are rethinking their economic relationship with artificial intelligence, trying to ensure their communities stand to benefit more clearly from a technology that may mint Silicon Valley an incalculable new fortune.
With A.I., the nation’s largest tech companies are seeking to revolutionize almost every field imaginable, from medical research to warfare. But, among Americans, there is growing fear that the financial gains of any breakthroughs may not be shared — or, worse, that they may come at their direct expense.
Jobs could be lost. Industries could be destroyed. Natural resources could be drained. Entire communities could be reshaped by A.I., with little guarantee that the public would be protected.
The disparity has triggered new debate in Washington about if, or how, the government could share in the wealth of A.I. companies and redistribute their profits. Among state and county officials, meanwhile, there is a new push to tax or otherwise regulate the industry. And hundreds of communities, led by Democrats and Republicans alike, have already tried to ban the construction of data centers, at least temporarily, as they sort through the economic implications.
A.I. companies have argued they can work with government to contain job losses or other downsides. But their generosity has its limits, and the political groups that represent Big Tech have lobbied against proposals they see as too demanding — setting up clashes around the country with communities that believe they are owed more.
The result has transformed A.I. into one of the most salient political issues in the 2026 midterm elections, and the outcome could shape the trajectory of the U.S. economy for decades to come.
“We should be asking, what does A.I. owe us?” said Betsey Stevenson, a professor of public policy at the University of Michigan and a chief economist at the Labor Department during the Obama administration. “What is my rightful inheritance from A.I.?”
At the heart of the A.I. policy debate are two possibilities.
The first is that A.I. will create an unfathomable economic boom, perhaps adding more than $15 trillion to the world’s economic output by 2030. The forecast alone has supercharged the valuations of major tech companies, which now represent a lopsided proportion of the S&P 500. Two young A.I. firms, Anthropic and OpenAI, are considering public stock offerings that may value both at $1 trillion or more.
The second possibility is a doomsday brought about by that same pursuit of growth. If A.I. falters, it could cause a catastrophe for markets, while if A.I. succeeds, it could leave millions of Americans unemployed. And it could concentrate more wealth in the hands of a few, like Elon Musk, who briefly became the world’s first trillionaire after one of his companies, SpaceX, went public in June.
The two paths have started a new chapter in the long-running political debate about inequality in America, where major economic transformations — from electrification to the deluge of cheap Chinese exports — have parceled out their costs and benefits disproportionately. The scars of that history are evident in countless towns that saw factories and storefronts darken as the economy globalized.
With A.I., the effects of automation largely have not yet been felt. There are no mass layoffs, for example, with robots replacing humans at scale. But the fear alone has helped provoke countless protests against the technology at all levels of government, even surfacing at one of President Trump’s rallies last month.
“The political pressure is growing because people are worried,” said Brent Orrell, the co-director of a joint A.I. commission started by the American Enterprise Institute, a conservative-leaning group, and the Urban Institute, a left-leaning group. “They’re worried about their economic well-being, and they see this as another threat.”
The concerns led roughly 200 economists, industry leaders and former regulators to issue a warning in July. Led by Erik Brynjolfsson, a Stanford economist and A.I. expert, their joint statement — “We Must Act Now” — urged governments everywhere to “build the incentives, guardrails, and institutions needed to steer A.I. in a direction that complements humans and benefits society.”
That conversation is only just beginning in hyperpartisan Washington, where A.I. carries high stakes for a country that derives much of its tax revenue from workers’ and businesses’ incomes. It has stoked another timeless debate about the role of the federal government — the balance it should strike as a facilitator of A.I. growth, and as a caretaker for those who may be harmed.
“A.I. and robotics are the most transformative technologies by far in the history of humanity,” Senator Bernie Sanders, the Vermont independent, said in an interview. “They have an enormous impact on our economy and the possibility that many millions of jobs that will be lost.”
Already, some in Congress have tossed around a battery of ideas, including new taxes that would apply to A.I. companies’ services and calls for a rethink of corporate rates. Particularly among Democrats, there is a desire to capture more money for the federal budget. The United States has a debt approaching $40 trillion, and it might grow further if Washington is forced to provide a financial cushion for those displaced by A.I.
Mr. Sanders has proposed that the U.S. government try to offset those consequences directly by taking a 50 percent stake in major A.I. companies. Under his plan, the United States would impose a one-time tax on tech giants, paid in equity to a newly established sovereign wealth fund. That fund would be worth about $7 trillion, based on current valuations. With the money, the government could immediately make a payment of more than $1,000 to all Americans, Mr. Sanders estimates, and then continue to redistribute any additional wealth as it is generated.
The idea is not without precedent in the United States. For decades, Alaska has paid small dividends to its residents based on the sale of oil extracted within its borders. The amount is hardly enough to replace a person’s wages, but for some economists, the fund is a tempting model for the digital age. If data is the oil for the nation’s A.I. giants, supporters reason, then the wealth they stand to generate from the world’s information should be shared in kind.
Mr. Sanders is not the only one floating such an option. Even his ideological foe, Mr. Trump, has mused about the public owning a portion of OpenAI. While the president generally has supported the pace of A.I. development — and castigated those who have stood in the way — he still signaled this year that he hoped to convene tech executives to discuss whether to take stakes in their companies.
In theory, some industry leaders, including Sam Altman, the chief executive of OpenAI, and Dario Amodei, the chief executive of Anthropic, support policies that redistribute A.I. wealth to citizens. Mr. Altman has even lobbied Mr. Trump on the matter. Others, like Mr. Musk, have similarly called for “Universal HIGH INCOME via checks issued by the Federal government.”
To date, though, none of the major tech companies has endorsed specific state or federal legislation, including the proposal from Mr. Sanders. Instead, they’ve primarily opted to call for studies.
OpenAI and Anthropic declined to comment. (The New York Times has sued OpenAI and Microsoft, claiming copyright infringement of news content related to A.I. systems. The two companies have denied the suit’s claims.)
But some experts are not sold on the idea of the government taking a stake, which could end up benefiting the A.I. companies more than the people who are disrupted by technology. A government that stands to profit from A.I. giants could be less inclined to regulate those companies in a way that could undermine their value.
“The prospect for abuse is so vast that everyone should be uncomfortable,” said Ben Harris, the vice president and director of Economic Studies at the Brookings Institution.
He added that Washington already has methods at its disposal to harness the gains from A.I. “We have really good tools for taxing capital,” he said. “One option is just to raise the rates.”
Far beyond Washington, the communities that host the lifeblood of the industry — the sprawling, resource-intensive computing centers that make A.I. possible — have similarly started to ask if they, too, are owed a greater share of future wealth.
Some have reconsidered their tax codes, as they question whether their ailing budgets can afford to subsidize Silicon Valley at the same time that they are making cuts to social services. Others have introduced new fees on data centers to offset their effects on local energy grids.
And still other places have halted construction outright. Over the past two years, roughly 120 communities have considered or adopted data center moratoriums of varying lengths, according to a tracker compiled by the Digital Technology for Democracy Lab at the University of Virginia.
Among the states to act is Nebraska, where abundant land and friendly tax policies have spurred Alphabet and other companies to set up new data centers. But surging demand led Gov. Jim Pillen, a Republican, to halt a state law that had offered tax incentives to the industry, arguing in July that the “market has changed dramatically.”
“We don’t need Nebraskans to pay,” he said. “Big Tech needs to pay their way.”
At the heart of the policy response is a fundamental shift in the relationship between local government and powerful interests in Silicon Valley.
Once, communities nationwide eagerly courted companies like Amazon, Meta, Alphabet and Microsoft, lavishing them with tax incentives and other perks. Chasing an economic boom, local officials were willing to make short-term fiscal sacrifices in exchange for coveted jobs and potential tax revenue.
But the industry’s demands — and profits — grew exponentially. Sometimes, their promises did not pan out. And the new data centers that they erected to serve A.I. have carried a much different footprint than earlier facilities. The new warehouses are larger, employ fewer people, and can consume far more water and electricity, stoking local fears about their economic and environmental impact.
New York State has taken some of the most aggressive steps in response. In July, Kathy Hochul became the first governor to enact a statewide moratorium on data center development for the next year. She also promised a wholesale reconsideration of the state’s tax treatment of data centers.
In an interview, Ms. Hochul stressed that she was not opposed to the tech industry, citing her state’s previous efforts to attract a $100 billion investment from Micron, a semiconductor manufacturer, in the hopes that it would create thousands of jobs in upstate New York.
Ms. Hochul argued that the rapid construction of A.I. data centers was categorically different. She said they required more power than the state could provide, even if it could quickly construct a new nuclear plant. Nor, she added, did the data centers appear likely to create many long-term jobs; they could even have the opposite effect.
“What’s the return on investment,” Ms. Hochul said, “when the jobs aren’t associated with it?”
Many in the tech industry have criticized New York and similar moratoriums considered or adopted elsewhere. The opponents include the Data Center Coalition, a lobbying group representing major tech companies including Amazon, Microsoft and Alphabet, as well as lesser-known data center developers.
Dan Diorio, the executive vice president of state policy and government affairs at the organization, said that the New York policy was too “open ended.” The bans, he added, also left some areas “out to dry with such a blanket policy that doesn’t actually account for the fact they want this.”
Some regions have indeed benefited from data centers, which have brought a rush of construction jobs, a glint of economic revival and, in some cases, a wellspring of new property tax revenue.
That includes Louisiana, where one parish was able to give bonuses to public teachers thanks to new revenue from a data center project. In Loudoun County, Va., the area known as Data Center Alley has generated so much economic activity and revenue that officials cut residents’ property taxes.
In a bid to capture similar economic benefits, other local governments have explored tough laws that would force tech companies and data center developers to pay up. But many of their proposals have drawn the sharp and well-funded opposition of an industry that now counts as one of the most powerful influences in American politics.
“What they usually say is, oh, of course, they would love to do that, but they don’t want to be compelled to do that,” said Senator Josh Hawley, a Republican from Missouri, who has long sought to rein in the industry.
The A.I. companies have been especially adversarial when policymakers have targeted their taxes. Broadly, tech giants have taken advantage of laws that allow them to buy computers, servers and other equipment without paying sales taxes.
When officials in Virginia tried to repeal that exemption — an idea proposed by Ms. Lucas, the state senator — the Data Center Coalition led a lobbying campaign to keep it intact. There, the group helped to create one of a series of dark money political organizations that have run ads nationally, seeking to defeat unfavorable tax policies or promote the benefits of the A.I. industry.
“Virginia’s data centers are there for you,” one of the ads began.
In the end, the coalition largely succeeded, working with others in the state to quash a repeal that Mr. Diorio later described as “punitive.” Instead, Virginia imposed a small, temporary and limited tax on the industry’s energy usage.
The outcome frustrated Ms. Lucas, who promised to keep fighting. “We’re missing those dollars that would have otherwise come to us,” she said.
Many of the states that have targeted data centers have encountered equally stiff opposition from Mr. Trump. Shortly after New York announced its moratorium, the president demanded that Ms. Hochul rescind the order and predicted that such bans would drive data centers to “red states.”
But the backlash has been bipartisan. This year, both Democratic-led Illinois and Republican-led Ohio have paused some tax incentives that benefit data centers. So has Arizona, which finalized a three-year pause in June that Katie Hobbs, the Democratic governor, said would help to free up money for child care and other social services.
In an interview, Ms. Hobbs said that while some people are “doing very well because of A.I.,” the benefits weren’t always widely shared. “And some of the pushback you’re seeing from communities is that a lot of people aren’t seeing that.”
Many cities and states have been spurred to action by voters who are concerned that they may be forced to foot the bill for the energy demands of data centers, after years of persistently high inflation.
In Texas, Gov. Greg Abbott, a Republican, once described his state as the “epicenter of A.I. development.” But this month Mr. Abbott ordered a statewide audit of data centers and their compliance with local electricity rules, part of a series of recent actions targeting the industry.
“Simply put, Texans must come first,” the governor said at the time.
Major companies including Anthropic, Meta, Alphabet and OpenAI have pledged to furnish more of their own power for data centers, so that they do not burden local residents with larger bills. Some have already done so, in accordance with a new White House pact meant to reduce costs.
But the industry’s efforts have varied in consequence — creating new energy demands in some communities, while reducing costs in others. Nor has progress come fast enough to satisfy all states. Over the past two years, roughly 30 states have considered legislation related to data center energy use, including special tariffs on the largest electricity users, including data centers, according to the University of Virginia tracker.
The list includes New Jersey, where lawmakers this year adopted a bill by Assemblyman David Bailey, a Democrat, who cited concerns about power costs as “one of the No. 1 sources of calls we were getting to our district office.”
The proposal angered the Data Center Coalition, which tried to circulate among legislators an edited version of the bill that would have exempted some facilities from the fees. But lawmakers still muscled through a version of the tariff, which Mr. Bailey said would “make sure the companies and developers were going to pay.”
“We’re learning from Ohio, we’re learning from Illinois, we’re learning from Virginia,” he said, of other states that had asserted their authority. “What are we really getting? What’s our return on investment here?”
The post With A.I. Riches at Stake, Pressures Mount to Share the Wealth appeared first on New York Times.
