Tech leaders and the Trump administration want you to believe that AI is the key to a new golden age. The reality looks more like a bold daylight heist.
Illustration by Adrià Fruitós.
Late in the afternoon of July 23, 2025, Donald Trump stood on a stage at the Andrew W. Mellon Auditorium in Washington, DC, to announce one of the hallmark initiatives of his second term as president: his AI Action Plan. Immediately after he took office, Trump had declared his administration’s intention “to sustain and enhance America’s global AI dominance in order to promote human flourishing, economic competitiveness, and national security.” Now, after weeks of consultation with stakeholders, he was ready to unveil his plan to a room filled with corporate leaders eager to see whether he would deliver for them.
“From this day forward, it’ll be a policy of the United States to do whatever it takes to lead the world in artificial intelligence,” Trump promised, flanked by a pair of large signs reading “Winning the AI Race.”
The choice of venue was fitting. Andrew Mellon was a powerful industrialist and banker who served as secretary of the treasury during the economic boom of the 1920s and through the Wall Street Crash of 1929. His fiercely pro-business, anti-tax policies are widely blamed for creating the conditions that led to the Great Depression. Whether AI will have the same effect on the economy is the central question for policymakers encountering the heady excitement and anxiety swirling around this new technology.
Trump, however, didn’t seem to harbor such concerns. Indeed, the three “pillars” of his AI Action Plan make clear that his administration intends to tip the scales in favor of the industry’s interests in a manner unprecedented in US policymaking. These pillars are: a push to mobilize “every tool at our disposal to ensure that the United States can build and maintain the largest, most powerful, and most advanced AI infrastructure anywhere on the planet”; a commitment to “get the entire world running on the backbone of American technology” by mobilizing government resources behind a global sales pitch on behalf of AI companies; and a determination that the government would divest itself of any use of “woke” AI models.
The industry reps in the audience were thrilled. But if any members of the general American public had been in the room, they might have wondered: What about us? As our resources—our land, tax dollars, jobs, and future—are handed to an industry that is far more interested in amassing money and hoarding control than democratizing them, what can we expect in return? Are we really witnessing the dawn of a new “golden age,” as Trump promised? Or, rather, a brazen daylight heist?
The billionaires selling us AI technologies would have us believe that they are self-made innovators who’ve built the most promising industry of our time based solely on their brilliance and entrepreneurial spirit, but the reality is far less valiant. For one thing, the paradigm of large-scale AI is characterized much more by brute-force resource consumption (of data, energy, and the capital that powers these infrastructures) than by scientific advancement. As Meredith Whittaker, the president of Signal (and a cofounder of AI Now), has observed, “It was not the algorithm that was a breakthrough: It was what the algorithm could do when matched with large-scale data and computational resources.”
At the same time, the choice to orient around the notion that “bigger is better” means that the AI industry is trapped in a business paradigm that depends on access to unfathomably large amounts of capital to build out its infrastructure at a scale far removed from the actual indicators of demand, let alone any convincing signals of business viability. And enduring such stratospheric levels of uncertainty and risk requires nothing short of a cult-like belief that the industry will eventually prove economically transformative enough to justify these bets by a guarantor that can persuasively underwrite the market. It requires, in other words, underwriting at a scale that only the US government could meaningfully provide.
The Trump administration has stepped up to the challenge. It has not been shy about its use of the power of the pen to back the industry’s interests, from brokering sales deals with other countries on behalf of Nvidia, to backing a $1 billion loan to bring the Three Mile Island nuclear plant back online to power Microsoft’s AI data centers. And it has done all this despite the swelling opposition within both its MAGA base and the general public, who are growing uncomfortable with a technology that is being used to endanger people’s livelihoods.
The Trump administration’s AI policy is being led by its artificial-intelligence and cryptocurrency czar, David Sacks, who is a prolific investor in the AI industry, and Michael Kratsios, a former executive at Scale AI who now heads the Office of Science and Technology Policy. Sacks and Kratsios have championed a multipronged approach that includes the $1 billion in AI funding provided by the One Big Beautiful Bill Act and an aggressive export agenda that turns the government into the top-level salesman for AI firms as they enter foreign markets. At the India AI Impact Summit in February, Kratsios announced the formation of a new Tech Corps, which will leverage the infrastructure of the Peace Corps to send technologists around the world on behalf of US tech firms to assist governments in integrating the companies’ software into their public-service systems. A few months earlier, the Department of Energy announced its “Genesis Mission,” a set of “private-public partnerships” through which the DOE will give companies access to its highly prized genomic and other datasets to enable them to develop products for commercial use. This is all on top of the already heavy subsidies that data centers receive, including significant state and local tax breaks and federal subsidies for factory construction.
According to the big AI companies, this kind of ambitious and unconditional government support is just what’s needed to achieve their aim of limitless AI infrastructure expansion—which they assert will be necessary to reach the holy grail of artificial general intelligence, and to do so before China does. Under this arms-race logic, any restraint on corporate power is recast as an impediment to national-security interests and plainly unpatriotic—like blocking the Apollo program or the Manhattan Project (both of which, AI boosters insist, are worthy historical analogies).
But if there’s one thing we should have learned from past eras of technological transformation, it’s that the promotion of national monopolies does not necessarily lead to national competitiveness. Nor does it lead, seamlessly, to sustainable jobs, enduring employment, wage growth, and innovation. While it can lead to great wealth for some, it rarely guarantees the kind of mass national renewal that the tech elite and their friends in government promise.
Even before February 28, the reasons for Donald Trump’s imploding approval rating were abundantly clear: untrammeled corruption and personal enrichment to the tune of billions of dollars during an affordability crisis, a foreign policy guided only by his own derelict sense of morality, and the deployment of a murderous campaign of occupation, detention, and deportation on American streets.
Now an undeclared, unauthorized, unpopular, and unconstitutional war of aggression against Iran has spread like wildfire through the region and into Europe. A new “forever war”—with an ever-increasing likelihood of American troops on the ground—may very well be upon us.
As we’ve seen over and over, this administration uses lies, misdirection, and attempts to flood the zone to justify its abuses of power at home and abroad. Just as Trump, Marco Rubio, and Pete Hegseth offer erratic and contradictory rationales for the attacks on Iran, the administration is also spreading the lie that the upcoming midterm elections are under threat from noncitizens on voter rolls. When these lies go unchecked, they become the basis for further authoritarian encroachment and war.
In these dark times, independent journalism is uniquely able to uncover the falsehoods that threaten our republic—and civilians around the world—and shine a bright light on the truth.
The Nation’s experienced team of writers, editors, and fact-checkers understands the scale of what we’re up against and the urgency with which we have to act. That’s why we’re publishing critical reporting and analysis of the war on Iran, ICE violence at home, new forms of voter suppression emerging in the courts, and much more.
But this journalism is possible only with your support.
This March, The Nation needs to raise $50,000 to ensure that we have the resources for reporting and analysis that sets the record straight and empowers people of conscience to organize. Will you donate today?
Given the headiness of the moment we’re in, it’s easy to forget the lessons of history: those technological paradigms that upended the economic status quo before AI.
Take semiconductors. Originally a small part of the larger defense industry—semiconductor manufacturing was initially funded and planned in close coordination with the military, and semiconductors were purchased almost exclusively by the military-industrial complex—they did not significantly penetrate the civilian market until the late 1970s. Around that time, however, Japan’s production of high-tech goods began to eclipse the United States’ productive capacity, offering superior quality at cheaper prices, leading US semiconductor executives to realize that this threatened their own industry. Producers like Intel and Fairchild Semiconductor flooded Washington with lobbyists, begging the new Reagan administration for assistance and warning that if they didn’t get it, the effects on the US economy and national security would be dire.
What made these requests especially notable wasn’t just the ask for help—the aerospace and electronics industries also received government support—but the scale of the aid sought by these famously libertarian “Semiconductor Cowboys.” What also made their requests notable was that the White House listened.
From Ronald Reagan to George H.W. Bush to Bill Clinton, successive administrations offered significant assistance to semiconductor companies, ranging from facilitating opportunities for US firms to learn the superior Japanese production processes, to providing subsidies for new factory construction and antitrust exemptions, to forging coercive trade deals that favored US firms and processes. In 1986, for example, the Reagan administration used the threat of economic sanctions to secure 20 percent of the Japanese market for US producers. In the mid-1990s, the Clinton administration negotiated new international trade deals that made these US subsidies legal while outlawing those favored by Europe and Japan.
The level and the one-sidedness of the support that the industry received was precedent-setting in ways that we should pay close attention to for what they might portend with AI. The economist Laura Tyson has pointed out that these trade deals forever changed US trade policy by prioritizing market access for companies over the protection of domestic jobs. It was, as Tyson and David B. Yoffie wrote in “Semiconductors: From Manipulated to Managed Trade,” “the first major U.S. trade agreement in a high-technology, strategic industry and the first motivated by concerns about the loss of high-tech competitiveness rather than concerns about employment. It was the first U.S. trade agreement dedicated to improving market access abroad rather than restricting market access at home.”
The reason that three consecutive administrations, across both parties, made these moves was that they aligned with the elite belief that access to leading-edge semiconductors figured at the heart of US military and economic primacy. But these forms of extreme subsidy and favoritism also had significant costs, both for the public good and for American security at large. Coupled with a lack of oversight, the government’s permissiveness toward antitrust violations allowed companies like Intel to monopolize the field. This led ultimately to stagnation, which degraded rather than enhanced innovation and competitiveness on the global stage, while also bearing costs to the public.
Today, semiconductor firms continue to require massive capital investment and provide low profit margins—all while receiving significant government support. Semiconductor plants, called “fabs,” are increasingly automated. The industry’s poor workplace conditions (including exposing workers to toxic chemicals), high turnover rates, and aging workforce have led to a worker shortage. Intel has failed to invest adequately in R&D and has previously spent its government subsidies poorly.
Notably, Intel is the same firm that the Trump administration took a 10 percent equity stake in last year in an effort to shore up the company’s finances and further bind it to the government. Because it failed to keep pace with the rest of the industry, Intel became vulnerable to such novel state measures, perpetuating the cycle of government intervention and neglect.
If the history of semiconductors offers a warning, so far we don’t appear to be listening. AI firms are receiving even less oversight than the semiconductor companies did, making them less likely to be accountable even as these firms create technologies that play a more central role in our lives as core infrastructure. If anything, the government has worked aggressively to deregulate a tech industry already enjoying laissez-faire treatment. The White House is currently pushing for the preemption of state regulations of AI—an effort that effectively revives the reviled “AI moratorium,” which would have banned for a decade the ability of states to enforce their own laws on AI firms. (That bill was shut down in the Senate last year by a vote of 99 to 1.)
Get unlimited access: $9.50 for six months.
The AI industry is also at a much more nascent stage in terms of its ability to deliver on its proof of concept. The strategic importance of semiconductors to the nation’s economic and military strength was already well proven before the Reagan administration weighed in. While the transformative potential of AI tools is clear, it is much farther away by comparison: The leaders of AI firms have described their tools as not yet reliable enough for military use and are struggling to establish their enterprise business lines.
The core justification for why AI is a national strategic priority—the potential development of artificial general intelligence—also remains speculative. A recent survey by the Association for the Advancement of Artificial Intelligence found that 76 percent of AI experts said they are dubious that AGI will be achieved under the current paradigm, which posits that AI technologies improve with increased data and computing power. If this scaling law fails to hold, it will strike at the heart of the industry’s case for ever-increasing capacity and sow real doubts concerning the fantastical levels of projected demand undergirding the current push for infrastructure expansion.
It’s worth reminding ourselves of what will happen for Americans if this technology does succeed. If the vision for AI ends in mass job displacement, which is certainly what the explosive revenue projections of these firms are banking on, then the administration would be inviting a new crisis onto its doorstep. Beyond job loss, the potential harms of AI are legion, including higher energy costs, negative effects on education, increased social isolation, environmental degradation, and more. All will likely require substantial and costly government intervention.
Already, according to the latest report on US manufacturing, the construction of data centers is crowding out the development of other industrial sectors. “Global logistics remains sensitive to geopolitical shifts,” warned the Institute for Supply Management’s December 2025 Manufacturing PMI report, which also noted the negative effects of Trump’s tariffs. “Large-scale data center programs are absorbing and reducing availability of resources for other sectors.”
The opportunity costs of prioritizing the needs of this sector above all others are massive and will have ripple effects on the economy for decades, delaying the development of other sectors as well as national development as a whole. Even worse, all this sacrifice may be for very little payoff. We are trading the health of other manufacturing sectors for a bet on a speculative promise that AI will be worth the economic pain.
Given the clear harms incurred by such a haphazard, speculative strategy, it’s worth asking why we are betting everything on this particular technology. These movements may not be reflective of an industrial-policy moon shot so much as old-school patronage and crony capitalism: Witness David Sacks’s brokering of access to the White House for his industry colleagues, recommending policies that benefit them, and retaining stakes in nearly 450 companies that would be aided by his policies.
It isn’t hard to see how Sacks and the tech CEOs he considers colleagues will benefit from these moves. But it’s less clear what American taxpayers will get in return, and this is roiling Trump’s MAGA base. “It feels like millions of votes across the country just got traded for thousands of [venture capitalist] and tech rich votes in regions Republicans will never win,” a conservative supporter of the administration told The Washington Post recently.
To make such a huge and risky bet on a technology that promises—even according to its greatest prophets—to fundamentally reshape the economy in ways that could put millions of people out of work and further centralize power and wealth in Silicon Valley certainly reads like a betrayal of that original vision. We should hope that any American government would support the conditions for beneficial innovation, but not wildly gamble away our future for so little in return.
Susannah GlickmanSusannah Glickman is an assistant professor at Stony Brook University.
Amba KakAmba Kak is, with Sarah Myers West, co–executive director of AI Now, an independent research institute.
Sarah Myers WestSarah Myers West is, with Amba Kak, co–executive director of AI Now, an independent research institute.