Meta, Microsoft, Amazon and others using long-term nuclear PPAs and bespoke projects to power AI data centers

The hyperscale AI sector’s strategic reliance on nuclear power to underpin carbon-neutral, high-density energy supply continues to deepen and evolve through 2026, with major players—Meta, Microsoft, Amazon, and Apple—expanding their vertical integration and pioneering bespoke nuclear–data-center collaborations. These developments coincide with emerging supply chain tightness, technological advances, and shifting policy landscapes, underscoring nuclear energy’s indispensable role in sustaining AI’s exponential growth while exposing critical vulnerabilities that demand urgent, coordinated action.

---

Hyperscale AI Firms Accelerate Vertical Nuclear Integration and Bespoke Campus Deployments

Building on prior commitments, Meta, Microsoft, Amazon, and Apple have intensified their involvement across the entire nuclear fuel cycle—from long-term PPAs to upstream uranium sourcing, HALEU fuel partnerships, and advanced reactor collaborations tailored specifically for AI data centers’ 24/7, high-density power demands.

• Meta now commands over 2,600 MW of nuclear PPAs and is advancing a 1.2 GW advanced nuclear campus in Ohio with Oklo Inc., embodying a tightly integrated co-location strategy that synchronizes nuclear generation with AI compute infrastructure. This approach reduces transmission losses and mitigates grid stress but has drawn scrutiny amid public concerns over transparency following reports like “BLACKOUT INCOMING - Meta’s Secret Nuclear Deal.”

• Microsoft has invested over $50 billion in nuclear-powered AI infrastructure, securing direct contracts for HALEU fuel and collaborating with innovators like Helion Energy. Its partnership with the Department of Energy (DOE) to scale domestic enrichment and fuel fabrication capabilities highlights a concerted effort to diversify supply and mitigate geopolitical risks.

• Amazon achieved a critical upstream milestone as its supported HALEU fuel fabrication facility, TRISO-X, recently obtained the Nuclear Regulatory Commission (NRC) Special Nuclear Material license, the first such licensing in the U.S. in over 50 years. This facility is positioned to alleviate HALEU fabrication bottlenecks, directly supporting Amazon’s expanding AI data center fleet.

• Apple has strategically secured upstream uranium supply through a partnership with Canadian miner NexGen Energy, providing a buffer against geopolitical volatility and recent U.S.–Canada uranium trade tensions.

Complementing these moves, bespoke nuclear–data-center campus partnerships are proliferating. The Constellation–CyrusOne collaboration in Texas exemplifies this trend by integrating advanced nuclear reactors with hyperscale AI compute on shared campuses, maximizing operational efficiencies and managing grid constraints effectively.

---

Supply Chain Stress Points: Uranium Mining, HALEU Enrichment, Fabrication, and Geopolitics

The supply chain underpinning the nuclear-powered AI surge faces mounting pressures:

• NexGen’s Rook I uranium mine in Canada represents a pivotal new source but comes with high production costs that anchor uranium prices upward, exacerbating supply tightness. This mine, approved recently after two decades without new Canadian uranium projects, signals both opportunity and risk amid geopolitical frictions, particularly ongoing U.S.–Canada uranium trade disputes threatening cross-border supply stability.

• HALEU enrichment capacity remains a critical bottleneck. Despite federal initiatives and projects like Centrus Energy’s Project Vault in Ohio and Orano’s U.S. enrichment filings, experts warn existing and planned capacities will fall short of meeting a decade-long surge in demand fueled by AI data centers and advanced reactors.

• The TRISO-X licensing milestone partially alleviates HALEU fabrication constraints but does not fully resolve enrichment shortfalls, highlighting a persistent pinch point in the fuel supply chain.

• Global geopolitical dynamics intensify risks: China’s consolidation of uranium supplies could reduce Western uranium availability by up to 36% by 2040, underscoring the urgency for Western nations to expand indigenous uranium enrichment and fabrication capabilities.

• Innovative firms such as ASP Isotopes and Quantum Leap Energy are advancing novel HALEU production technologies, while companies like Global Atomic explore alternative uranium extraction methods, including zinc recycling, as the industry pursues diversified, resilient supply chains.

---

Emerging Technological Advances and Research Partnerships Bolster Nuclear Innovation

Recent breakthroughs and collaborative efforts aim to enhance the performance and efficiency of next-generation nuclear reactors powering AI infrastructure:

• The Texas A&M Engineering Experiment Station and ZettaJoule signed a landmark Memorandum of Understanding (MOU) to explore building a transformative Very-High-Temperature Gas Reactor (VHTGR). This reactor design promises higher thermal efficiencies and improved integration with AI data center energy demands.

• At the Department of Energy’s Argonne National Laboratory, researchers have scaled nuclear material testing to 1,340°F, advancing next-generation reactor materials and efficiency. These breakthroughs are critical to developing reactors capable of sustained high-output operations tailored to AI workloads.

• However, a recent investigative report warned of a looming fuel crisis that could stall the U.S. nuclear energy comeback, emphasizing that fuel supply chain constraints—particularly in enrichment and fabrication—pose a significant risk to scaling nuclear-powered AI infrastructure.

---

Policy, Regulatory, and Grid Modernization: Navigating Complexity Amid Accelerated Demand

Policy and regulatory landscapes continue to evolve rapidly in response to nuclear power’s growing role in AI infrastructure:

• State-level pro-nuclear initiatives are gaining momentum. For instance, Illinois Governor JB Pritzker’s Executive Order 2026-03 reinforces state support for new nuclear deployments and infrastructure upgrades serving data center clusters. Similarly, rural states like Utah actively seek to host nuclear projects, aiming for economic revitalization and job creation aligned with national nuclear expansion goals.

• At the federal level, the White House demands AI firms internalize the full costs of their energy consumption, advocating transparent cost allocation, grid impact mitigation, and sustainability integration. The Senate and DOE are simultaneously pursuing permitting reforms to fast-track advanced reactor commercialization and compress historically lengthy deployment timelines.

• Heightened media scrutiny and public pressure following exposés on opaque nuclear agreements have intensified calls for transparent, collaborative grid planning to prevent systemic risks like blackouts amid surging AI electrical loads.

• Utilities such as Duke Energy are investing billions in grid modernization—embracing smart grid technologies, virtual power plants (VPPs), and transmission upgrades—particularly in AI-intensive regions like North Carolina. Engineering firms like Quanta Services are executing critical transmission and substation projects that support reliable, large-scale nuclear power delivery.

• Globally, Australia’s APA Group is advancing a $3 billion transmission portfolio explicitly designed to accommodate AI load surges, signaling private-sector alignment with nuclear-powered AI growth worldwide.

• State regulators in New York and Washington are pioneering equitable cost-sharing frameworks and community engagement processes to harmonize infrastructure growth with environmental and social priorities.

---

Market Signals, Investor Sentiment, and Industry Warnings

The nuclear-AI nexus is eliciting mixed financial market responses:

• Centrus Energy’s CEO recently warned of a looming “supply gap” in enrichment capacity that could derail nuclear reactor expansions vital for AI data centers, urging accelerated investment and regulatory streamlining.

• Small Modular Reactor (SMR) stocks, including NuScale, have experienced market corrections amid slower-than-expected commercial traction, while capital markets remain active with transactions such as Deep Fission Nuclear’s HALEU supply agreement with Urenco USA and the Eagle Nuclear Energy SPAC merger targeting AI power needs.

• Notably, the Nasdaq exchange launched a dedicated nuclear market listing platform targeting hyperscale AI companies, signaling growing investor recognition of nuclear’s strategic importance despite persistent capital intensity and regulatory hurdles.

---

Persistent Supply Chain Tightness Reinforces Domestic Capacity Imperatives

• Uranium spot prices remain firm, stabilizing between $90–$100 per pound, reflecting ongoing production shortfalls amid surging demand.

• While new Canadian mines and U.S. exploration projects aim to expand supply, full materialization is years away, compounding near-term tightness.

• HALEU fabrication capacity is improving with TRISO-X’s licensing, but enrichment remains a critical bottleneck threatening downstream fuel availability and reactor deployment timelines.

• Innovative HALEU production technologies and alternative uranium extraction approaches offer promise but require accelerated scaling and investment.

• Intensifying geopolitical frictions—especially China’s uranium supply consolidation and U.S.–Canada trade tensions—underscore the critical necessity of securing and scaling indigenous enrichment and fabrication infrastructure to safeguard supply chain resilience.

---

Conclusion: Urgent, Coordinated Action Is Essential to Sustain Nuclear’s Role as AI’s Carbon-Neutral Backbone

As 2026 progresses, nuclear power’s position as the cornerstone of a sustainable, carbon-neutral AI economy is unequivocal. Hyperscale AI firms’ expanding vertical integration—from uranium mining to enrichment, fabrication, and integrated nuclear–data-center campuses—cements nuclear energy as indispensable to powering AI infrastructure at scale.

Yet, this strategic pivot faces multifaceted challenges:

• Critical enrichment and fabrication bottlenecks risk throttling AI data center expansion.

• Geopolitical supply risks and trade tensions threaten supply chain continuity.

• Rising political and public demands for transparency, grid reliability, and equitable cost-sharing add pressure for responsible deployment.

• Modernizing transmission and distribution infrastructure to handle rising AI loads is imperative.

Addressing these challenges demands:

• Accelerated domestic enrichment and fabrication capacity expansions

• Robust grid modernization and transmission investments

• Regulatory agility with transparent, collaborative planning frameworks

• Inclusive community engagement and equitable cost-sharing mechanisms

Only through decisive, coordinated efforts by industry, policymakers, regulators, utilities, and communities can nuclear energy reliably power the AI economy’s explosive growth—ensuring a resilient, carbon-neutral future for hyperscale AI and beyond.