How America and Israel Are Building the Infrastructure of the Future
In December 1960, when American U-2 surveillance planes first photographed the construction site in the Negev Desert, Israeli officials told the Eisenhower administration they were building a textile factory. The textile factory was Dimona—Israel’s nuclear weapons facility, which would eventually produce an estimated 80 to 400 nuclear warheads, depending on whose intelligence you believe. That original act of strategic deception created a legal paradox that persists to this day. Israel has never signed the Nuclear Non-Proliferation Treaty. It has never admitted to possessing nuclear weapons. It maintains a policy of “nuclear ambiguity”—neither confirming nor denying what everyone knows to be true. This studied silence has served Israel well for six decades, but it also created a constraint: under American law, the United States cannot transfer nuclear technology to non-NPT signatories.
So when the Trump administration announced plans to help build a civilian nuclear reactor on Israeli soil, a simple question emerged: How?
The answer reveals something important about how American power actually works—not through treaty revision or congressional authorization, but through jurisdictional creativity.
The proposed reactor will sit on Israeli territory but operate under American legal authority. This is not a euphemism. The facility will function, for purposes of U.S. nuclear law, as if it were located in Tennessee or South Carolina. American operators, American safety protocols, American jurisdiction. The Nuclear Regulatory Commission will oversee it. The fuel will remain American property. Israel provides the land and the electricity demand; America provides everything else—including the legal fiction that makes the arrangement possible. This is not “technology transfer” in any sense that the NPT’s architects contemplated. It is sovereignty engineering—the deliberate construction of overlapping legal jurisdictions to achieve what direct cooperation cannot. Critics will call it a loophole. Supporters will call it elegant. Both are correct.
To understand why Israel needs this reactor—and why America is willing to perform such legal contortions to provide it—you need to understand a single number: 200 megawatts. That is the approximate power consumption of a single large-scale AI data center. Not a cluster of centers. Not a campus. One facility. The equivalent of powering a city of 150,000 people—devoted entirely to running servers. Israel wants to become a major hub for artificial intelligence development. It has the talent: more engineers per capita than any nation on Earth, a startup ecosystem that produced Waze, Mobileye, and the core technology behind Intel’s autonomous driving division. Israeli chip designers are among the most sought-after in the industry. The human capital exists.
What doesn’t exist is energy.
Golda Meir’s famous quip has never been more relevant: “Moses dragged us through the desert for forty years to bring us to the one place in the Middle East without oil.” Israel discovered significant offshore natural gas reserves in 2010—the Tamar and Leviathan fields—but gas has limitations. It works for heating and electricity generation, but AI data centers require something more: baseload power that runs 24 hours a day, 365 days a year, without interruption and without carbon emissions that would trigger European regulatory penalties. Nuclear is the answer. Perhaps the only answer.
This explains the Mar-a-Lago meeting in December 2024, when Netanyahu and Trump discussed energy cooperation alongside the usual topics of Iran and regional security. The public readouts emphasized strategic alignment. The private conversation, according to officials familiar with the discussions, centered on a more prosaic question: How does Israel get the power it needs to matter in the AI race? The reactor deal answers that question. Israel gains energy independence—not just from oil it never had, but from the gas it does have, which can now be exported rather than consumed domestically. America gains something equally valuable: a trusted node in the emerging architecture of technological competition.
The phrase sounds grandiose, but the concept is straightforward. Just as Pax Romana described a Roman-dominated world order maintained through military supremacy, and Pax Americana described a U.S.-dominated order maintained through economic and military power, “Pax Silica” describes the emerging competition to control the infrastructure of the digital age. This infrastructure has four components: semiconductor manufacturing, AI development capacity, rare earth minerals, and energy supply. The nations that control these four pillars will shape the next century of human history. The nations that don’t will be shaped by it. American strategists understand this. The CHIPS Act, passed in 2022, allocated $52 billion to rebuild domestic semiconductor manufacturing. The Inflation Reduction Act directed billions more toward energy infrastructure and battery production. But legislation alone cannot secure American technological supremacy. America needs allies—trusted partners who can extend the supply chain beyond U.S. borders without creating vulnerabilities.
The emerging coalition is visible if you know where to look: Israel for chip design and AI software. Taiwan for advanced semiconductor manufacturing (though this node is dangerously exposed to Chinese pressure). Japan and South Korea for memory chips and manufacturing expertise. Singapore as a financial and logistics hub. The UAE as an energy provider and gateway to emerging markets. The United Kingdom for research and intelligence. This is not a formal alliance in the NATO sense. There is no treaty, no mutual defense obligation, no joint command structure. It is something more fluid: a network of bilateral arrangements, technology-sharing agreements, and investment relationships that together constitute a “trusted zone” for the infrastructure that matters most.
Israel’s position in this network depends on solving its energy problem. A tech hub without reliable power is not a hub at all—it’s a dependency. The reactor changes that calculation. It transforms Israel from a talent pool that can be poached into a self-sustaining node in the American technological architecture.
There is a reason why every major technology company is suddenly interested in nuclear power. Microsoft has signed agreements to restart Three Mile Island. Google has invested in small modular reactors. Amazon has purchased a nuclear-powered data center campus in Pennsylvania. OpenAI’s Sam Altman has invested heavily in nuclear startups. The pattern is unmistakable: the AI industry has realized that its future depends on solving a physics problem, not just an engineering one. Large language models like GPT-4 require immense computational resources—not just for training, which consumes electricity in massive bursts, but for inference, the ongoing process of responding to user queries. Each conversation with an AI chatbot consumes roughly ten times the electricity of a traditional Google search. Multiply that by hundreds of millions of users, and you begin to understand the scale of the problem.
Current AI data centers draw anywhere from 50 to 500 megawatts of power. The next generation of facilities—the ones needed to train models more powerful than anything that exists today—may require gigawatts. A single gigawatt is the output of a large nuclear power plant. It is more electricity than many small countries consume in total. Solar and wind cannot solve this problem. They are intermittent—dependent on weather, time of day, and season. AI training runs cannot pause when the sun sets or the wind dies. Battery storage helps at the margins but remains far too expensive to provide backup at the scale required. Natural gas works, but it produces carbon emissions that create regulatory and reputational risks, especially in Europe. Nuclear is the only proven technology that provides reliable, carbon-free baseload power at the scale AI demands. This is why the American technology industry has suddenly discovered a love for atomic energy. And it is why Israel’s reactor deal matters far beyond the Middle East.
Israel is positioning itself as a node in the global AI infrastructure—not just a source of talented engineers, but a location where computation can happen at scale. The reactor is not incidental to this strategy. It is foundational.
For decades, strategists spoke of the “resource curse”—the paradox that nations rich in oil or minerals often suffered from corruption, authoritarianism, and economic stagnation. The resource curse was a story about the pathologies of abundance. The new resource curse is about scarcity—specifically, the scarcity of the inputs that the technological economy requires. Advanced semiconductors require rare earth elements, ultra-pure water, and specialized manufacturing equipment. AI development requires chips, talent, and electricity. These resources are not evenly distributed. Taiwan manufactures over 90% of the world’s most advanced chips. China controls roughly 60% of rare earth processing. A handful of countries—the United States, China, Israel, the UK—produce the majority of AI research. This concentration creates vulnerability. A Chinese blockade of Taiwan would cripple global chip supply. A disruption in rare earth exports would halt electric vehicle and electronics production. A shortage of electricity would strand the most sophisticated AI models mid-training.
Nations are responding to these vulnerabilities by building redundancy and securing supply chains. The CHIPS Act funds new fabrication plants in Arizona and Ohio. European governments subsidize battery factories and chip foundries. Japan stockpiles rare earth elements and invests in recycling technologies. Israel’s reactor fits this pattern. It is an investment in resilience—a guarantee that Israeli AI development will not be held hostage to energy politics. Israel cannot control the price of natural gas or the geopolitics of pipeline routes. It can control a reactor on its own soil, operated under terms it has negotiated with its closest ally. This is the logic of technological sovereignty: the recognition that dependence on foreign inputs creates strategic vulnerability, and that true security requires controlling the infrastructure that makes advanced industry possible.
What connects these developments—the reactor deal, the technological alliances, the scramble for energy—is a fundamental transformation in what counts as power.
The first revolution is technological. For most of the twentieth century, military power meant industrial capacity—the ability to build tanks, planes, ships, and the logistics to deploy them. This remains important, but it is increasingly secondary. The primary arena of competition is now informational: who controls the algorithms, the data, the processing capacity. A nation that dominates AI development will enjoy advantages in every other domain—military, economic, cultural, diplomatic. The reactor deal is not about electricity. It is about ensuring Israel’s place in the architecture of AI dominance.
The second revolution is energetic. The green transition is not primarily an environmental project—it is a geopolitical one. Nations that decarbonize successfully will escape the constraints of fossil fuel dependency. Nations that don’t will remain vulnerable to the whims of petrostates and pipeline politics. Nuclear power is the underappreciated solution: zero-carbon baseload electricity that runs regardless of weather, geology, or diplomatic relations with hydrocarbon suppliers. Israel’s reactor is a down payment on this future.
These two revolutions are connected. You cannot win the technological competition without solving the energy problem. You cannot solve the energy problem without technological innovation. The nations that understand this linkage—and build infrastructure accordingly—will lead the coming century. The nations that don’t will follow.
What emerges from this convergence?
Israel gains nuclear reactors, unshakeable American backing, and a guaranteed seat at the table of the technological future. Its energy deficit—the constraint that might have reduced it to a minor player in the AI race—will be solved within a decade. The Negev, once valued only for its strategic depth and its symbolic role in Zionist mythology, becomes the site of a new kind of national project: not making the desert bloom, but making it compute. America gains a secure, trustworthy node in its emerging technological alliance. Israeli talent in chip design, cybersecurity, and AI development will be harnessed to American strategic interests, without the vulnerabilities that attend relationships with more ambivalent allies. The reactor deal locks Israel into the American technological orbit for decades—a commitment far more binding than any treaty. The broader architecture takes shape: a network of allied nations, bound not by ideology or mutual defense commitments, but by shared infrastructure and interdependent supply chains. Call it Pax Silica, call it the “trusted network,” call it whatever you like. The substance is what matters: a coalition organized around the technologies that will define power in the twenty-first century.
This is not the “rules-based international order” that diplomats invoke at conferences. It is something harder and more pragmatic: an infrastructure of advantage, built by nations that understand what the coming competition requires. Control the chips, and you control the computing power. Control the energy, and you can scale the infrastructure. Secure both, and you have built something that cannot easily be taken away.
The reactor in the Negev is not just a power plant. It is a statement about the future—and a bet that Israel and America intend to own it.
