The states building the most data centers are building the 21st century's strategic infrastructure. The states building across all three layers — ground, orbit, and edge — are building the infrastructure that survives.

The sovereign compute orthodoxy runs in a straight line. Compute follows energy. Energy follows geography. Geography follows politics. Every major AI infrastructure strategy on Earth — the US hyperscaler buildout, the Gulf’s sovereign data center investments, India’s Digital India stack — operates within this binding logic. Training a frontier model consumes enough electricity to power a small city. Cooling a hyperscale cluster requires millions of gallons of water annually. Permitting a grid interconnection takes years. The physical layer is real, expensive, and slow, and it anchors the entire AI stack to specific coordinates on a map.

This framing is correct. It is also incomplete. The sovereign compute race as currently understood is a race to concentrate AI infrastructure in fixed, targetable locations — data centers with GPS coordinates, grid substations feeding 500 MW to a single campus, subsea cables landing at known points. The Iran-Gulf conflict demonstrated what this concentration means in practice: three AWS facilities struck, Hormuz shipping down 90–95%, and sovereign AI strategies built on the assumption of physical security suddenly facing the oldest problem in strategic infrastructure. Concentration is capability. Concentration is also vulnerability.

Two emerging vectors suggest that the geographic binding constraints of sovereign compute — while real for training — may not be permanent for inference, which is where AI meets the population. Call the current orthodoxy the geopolitics of the physical: the assumption that AI power is determined by who controls the most land, energy, and water in the most politically stable territory. The two vectors I’m going to describe represent something different — the arbitrage of the digital: the possibility that inference capacity can be distributed across domains where physical geography no longer binds. One points up. The other points out. Neither is a clean solution. Together, they redraw the map.

The first vector is orbital.

In January 2026, Axiom Space launched the first two orbital data center nodes to low Earth orbit. Weeks earlier, NVIDIA-backed Starcloud ran Google’s Gemma large language model on a satellite carrying an H100 GPU — the first time a frontier-class chip operated from orbit. NVIDIA followed at GTC with a full space computing platform: Vera Rubin modules, IGX Thor, and Jetson Orin, with Axiom, Kepler Communications, Planet Labs, and Starcloud as launch partners. Google’s Project Suncatcher envisions constellations of TPU-equipped satellites connected by free-space optical links, with prototype launches planned for early 2027. China’s Zhejiang Laboratory has already deployed a 12-satellite computing constellation carrying an 8-billion-parameter AI model, while Beijing-based BAIST plans a megawatt-scale orbital data center by 2035. Two weeks ago, Syntiant and Novi Space demonstrated real-time AI object detection on a commercial satellite — with model retraining in under 24 hours.

Space sidesteps the terrestrial binding problem almost entirely. Solar power in LEO is continuous and abundant — a solar panel in the right orbit produces up to eight times more energy than on Earth. There are no permitting delays, no water rights conflicts, no grid interconnection queues. Thermal dissipation, though, remains the ultimate engineering frontier. On Earth, data centers use convection and liquid cooling. In vacuum, the option is radiation, and radiating waste heat from a megawatt-scale cluster requires massive deployable radiator arrays. Small-scale inference modules like Jetson Orin handle this with passive radiation. The 5-gigawatt architectures Starcloud envisions will require radiator wings at an industrial scale — heat management as the “orbital water rights” of the next decade. The strategic logic extends further: if compute dispersal is a resilience strategy, the Moon’s south pole, and specifically the Peaks of Eternal Light, where near-continuous solar exposure meets permanently shadowed craters for thermal management, becomes a redundancy node. The Artemis core program’s infrastructure is about depth-of-field for critical systems.

Orbital compute also introduces a survivability profile that terrestrial infrastructure cannot match. Starlink demonstrated this in Ukraine: a distributed constellation of thousands of satellites has no single node to destroy. A 5,000-satellite compute constellation presents a fundamentally different target than three data centers in a desert. Russia’s development of nuclear anti-satellite capabilities — a weapon the US Department of Defense warned could render LEO unusable for an extended period — is the only theoretical counter. But an EMP that disables all satellites in a wide orbital zone would destroy Russian, Chinese, and commercial assets alongside the target, making it a mutual-destruction weapon, not a practical one.

The second vector runs in the opposite direction — not up to orbit, but outward to the edge of the network, to every device in every pocket.

There are 7.4 billion active smartphones in the world today. India alone accounts for 1.2 billion smartphone connections. Mobile neural processing units now deliver throughput approaching data-center GPUs from 2017. Billion-parameter language models run in real time on flagship devices. IDC forecasts that over 370 million GenAI-capable smartphones shipped in 2025 — 30% of all shipments — rising to over 70% by 2029. Deloitte estimates inference workloads now account for roughly two-thirds of all AI compute, up from one-third in 2023. The on-device AI market is projected to grow from $33 billion in 2026 to $157 billion by 2033.

The strategic significance of edge compute is not mainly about latency or privacy. It is about distributing the physical substrate of sovereign AI to a target set so large it becomes indestructible. Iran can strike three data centers. It cannot strike 1.2 billion smartphones. No adversary can. The compute does not live in a building with coordinates — it lives in every pocket, on every wrist, behind every pair of smart glasses. Each device is a micro-sovereign compute node. Individually trivial. Collectively, an AI infrastructure that is physically unconcentrable and therefore physically undefeatable. The parallel to virtual power plants is direct: just as distributed solar and batteries aggregate into grid-scale capacity without centralized generation, distributed NPUs aggregate into national-scale inference capacity without centralized data centers.

India’s position in this framing shifts dramatically. In the current orthodoxy, India’s Intelligence layer — models, training capacity, frontier research — is its weakest dimension. But in a distributed inference world, India’s 1.2 billion connected devices become the largest sovereign compute substrate on Earth. The critical question becomes who controls the model weights on those devices. Apple’s Neural Engine runs Apple Intelligence. Qualcomm’s NPUs run whatever Qualcomm certifies. Google controls Android’s ML runtime across 3.9 billion devices globally. Edge AI may not democratize compute sovereignty — it may transfer it from territorial states to platform companies, making Apple and Google de facto sovereign compute providers for billions of people. This is why nation-state foundation models — India’s domestic efforts, the UAE’s Falcon family, Saudi Arabia’s SDAIA programs — become more strategically important in an edge-first world. Whoever controls the weights controls the intelligence at the edge, regardless of who manufactured the silicon.

Neither vector is a replacement for terrestrial infrastructure.

Training frontier models still requires concentrated, high-power, liquid-cooled GPU clusters. The data center buildout is necessary and serious. But training is not where most AI value will ultimately be captured. Inference is. And inference is splitting into three domains: terrestrial data centers, orbital compute, and edge devices. The sovereign compute race is no longer one-dimensional.

This reframing matters most for geographically constrained states. Singapore, the UAE, Israel suddenly have alternative paths to sovereign AI capability. The Gulf’s centralized data center investments are essential for training. Adding orbital partnerships and edge model ecosystems creates a three-layer resilience architecture where the ground layer can be struck, the orbital layer requires weapons that don’t yet practically exist, and the edge layer cannot be struck at all. A sovereign AI strategy built across all three layers is not just more capable. It is more survivable. And survivability, as the past two months have demonstrated, is not a theoretical concern.

The sovereign compute map most strategists are drawing — color-coded by data center capacity and grid megawatts — is a map of training infrastructure. The inference map has two additional layers that appear on no territorial chart: orbit and the edge. The states that see this early will not just build the most AI infrastructure, but the AI infrastructure that endures.

A note on independence: All opinions shared in this newsletter are my own and do not reflect the views of dmg events, ADIPEC, or any affiliated organizations. This is personal analysis, not institutional positioning.

Sources

NVIDIA, “NVIDIA Launches Space Computing, Rocketing AI Into Orbit,” GTC 2026

Axiom Space, “Orbital Data Centers,” January 2026

CNBC, “Nvidia-backed Starcloud trains first AI model in space,” December 2025

Google Research, “Project Suncatcher: Exploring a space-based, scalable AI infrastructure system design”

China Daily, “Chinese tech firms race to build AI computing capabilities in space,” December 2025

SatNews, “Syntiant and Novi Space Demonstrate Low-Power AI Inference in Orbit,” March 2026

Arms Control Association, “U.S. Warns of New Russian ASAT Program,” March 2024

Atlantic Council, “Russian nuclear anti-satellite weapons,” February 2024

DataReportal, Digital 2026 Global Overview

Deloitte, “Why AI’s next phase will likely demand more computational power, not less,” TMT Predictions 2026

IDC, Worldwide Smartphone Market Forecast, August 2025

Coherent Market Insights, On-Device AI Market, 2026–2033

Future Today Strategy Group, Convergence Outlook 2026

Washington is addressing chip enforcement in two halves: location verification and remote access controls. The third gap — end-use verification at the workload level — remains open in U.S. legislation. G42 built an operational layer for it.

On March 26, the House Foreign Affairs Committee advanced the Chip Security Act, a bill requiring location verification mechanisms embedded directly into every advanced AI chip before export.

The legislation advanced one week after prosecutors unsealed the indictment of Super Micro’s co-founder for conspiring to divert $2.5 billion worth of Nvidia servers to China through Taiwan and Southeast Asian intermediaries. Since November 2025, the Department of Justice has brought at least five separate smuggling cases involving fake audit servers, shell companies in Florida, and routing through Malaysia and Thailand. The Chip Security Act addresses this specific problem: physical diversion of chips from authorized destinations. Ping-based location verification, using network latency triangulation across a dozen geographically dispersed servers, can confirm a chip’s country-level position with roughly 60-mile accuracy. Working prototypes already exist on Nvidia H100 chips. Nvidia has acknowledged its GPUs already transmit the telemetry data needed for this, and implementation could come through a firmware update.

The enforcement question, though, is what happens when verification detects a chip in the wrong country. The bill’s text explicitly prohibits kill switches: it “may not be construed as requiring any chip security mechanisms that may hinder the capability or functionality of a covered integrated circuit product, such as a kill switch or geofencing mechanism.”

The chip keeps running. The exporter is required to report the discrepancy to the Bureau of Industry and Security. After that, enforcement becomes a law enforcement matter — DOJ investigations, sanctions, diplomatic channels. The Senate version directs Commerce and Defense to study whether future mechanisms could “modify the functionality” of illicitly acquired chips, but that assessment is at least a year away from any mandate. For now, detection and reporting is the entire enforcement architecture. Spoofing individual chips through latency manipulation is possible; spoofing thousands of GPUs simultaneously in a data center cluster is substantially harder, though the Center for Cybersecurity Policy has warned that any embedded verification mechanism introduces new attack surfaces that hostile actors could exploit.

Location verification addresses physical diversion. It does not address a second, arguably larger gap: Chinese companies legally accessing advanced chips in third countries. In March, the Wall Street Journal reported that ByteDance is deploying 36,000 Nvidia Blackwell B200 GPUs through Aolani Cloud in Malaysia — a $2.5 billion cluster that will be among the largest commercial AI deployments in Southeast Asia. The chips stay in Malaysia. They are formally owned and operated by a Malaysian entity. Nvidia confirmed the arrangement complies with export controls, noting that “by design, the export rules allow clouds to be built and operated outside controlled countries.” Location verification would confirm these chips are exactly where they are supposed to be. The problem is that ByteDance’s engineers access the compute remotely from China. The House passed the Remote Access Security Act 369–22 in January 2026 to close this “cloud loophole” by extending export controls to cover remote access to controlled chips. It has not yet cleared the Senate, and enforcement — monitoring who logs into every GPU cluster globally — is substantially more complex than verifying where hardware sits.

A third gap remains fully unaddressed by either bill. The Chip Security Act verifies where chips are. The Remote Access Security Act, if passed, would regulate who accesses them remotely. Neither verifies what workloads are running on the chips or whether the entity operating them is the authorized end user in practice, not just on paper. A Malaysian cloud operator that leases capacity to a Chinese AI lab is legal under current rules, and neither bill’s enforcement mechanism reaches the workload layer. This is the structural opening that G42’s compliance architecture occupies. The Regulated Technology Environment handles chip-level compliance — the State Department described it as a “gold-standard framework for managing sensitive technologies” at the first U.S.-UAE AI Acceleration Partnership Working Group meeting on March 26. The April 6 State Department readout went further, describing the RTE as operating under “jointly agreed security and governance requirements”. Digital Embassies, launched at Davos in January, establish government-to-government legal constructs that define who has jurisdictional authority over AI workloads, regardless of where infrastructure is physically located. And crucially, Microsoft serves as the primary infrastructure partner for this framework, reinforcing the UAE’s position as a collaborative node for U.S. technology rather than a sovereign replacement for it.

Greenshield, the operational layer run by Core42, applies cryptographic controls that verify where workloads are processed and ensure they remain within authorized sovereign environments. Together, the three layers create an end-to-end compliance architecture that covers physical location, access control, and workload verification — the full stack that U.S. legislation, taken together, still leaves partially open.

The Vietnam deal shows the model already scaling. In February, G42 signed a $1 billion framework agreement with FPT Corporation and Viet Thai Group to deploy three hyperscale data centers with full sovereign cloud capacity. Vietnam gets AI infrastructure without negotiating chip allocations or export licenses with Washington directly — it buys into a G42-managed ecosystem that has already cleared those compliance hurdles across all three layers.

The Atlantic Council’s William Wechsler argued this month that the Gulf’s interdependence with the West extends far beyond energy, noting the region “will have a disproportionate role in the future of the global infrastructure for artificial intelligence, just as Taiwan today plays an outsize role in semiconductors.” Taiwan’s leverage is its position as the irreplaceable manufacturing node. The UAE is building the analogous position for compliance: the node through which verified sovereign AI access flows to the Global South.

The critical question for policymakers and infrastructure investors: if chip traceability becomes the global standard for hardware access — and bipartisan momentum suggests some version will — then verified compliance infrastructure replaces raw chip supply as the binding constraint on sovereign AI. The Semiconductor Industry Association and the Information Technology Industry Council are right that mandated tracking could push some buyers toward Chinese alternatives. That risk strengthens the case for a trusted intermediary that keeps nations inside the U.S. technology stack while providing sovereign control at the workload level. Five federal smuggling cases, a $2.5 billion legal cloud workaround in Malaysia, and an open gap at the end-use layer all point in the same direction. Whoever controls the compliance architecture that determines where, by whom, and for what purpose GPUs operate holds the next structural advantage in sovereign AI.

Sources

• H.R. 3447 / S. 1705 — Chip Security Act, 119th Congress (congress.gov)

• H.R. 2683 — Remote Access Security Act, passed House 369–22, January 12, 2026 (eeNews Europe)

• House Foreign Affairs Committee press release, March 26, 2026

• U.S. State Department — First U.S.-UAE AI Acceleration Partnership Working Group, April 6, 2026

• CAIS Action Fund — “The Chip Security Act: Separating Fact From Fiction”

• CSIS — “The Architecture of AI Leadership,” February 2026

• Center for Cybersecurity Policy & Law — analysis of CSA cybersecurity risks, July 2025

• Foundation for American Innovation — “The Chip Security Act: A Bipartisan Solution,” July 2025

• Wall Street Journal — ByteDance / Aolani Cloud / Blackwell deployment, March 13, 2026

• Tom’s Hardware — Nvidia confirmation: “export rules allow clouds outside controlled countries,” March 2026

• BISI — “AI Chip Smuggling: The Limits of US Export Controls,” April 6, 2026

• SIA Statement on Chip Security Act, March 2, 2026

• ITI — “The Unintended Consequence of the Chip Security Act,” February 27, 2026

• G42 — Digital Embassies and Greenshield launch, January 20, 2026 (Davos)

• G42 / FPT / Viet Thai Group Framework Agreement, February 9, 2026

• Atlantic Council — William F. Wechsler, “Why the Gulf Should Keep Looking West,” April 2026

A note on independence: All opinions shared in this newsletter are my own and do not reflect the views of dmg events, ADIPEC, or any affiliated organizations. This is personal analysis, not institutional positioning.

Eighteen months ago, utilities sold electricity and hyperscalers bought it. That boundary is dissolving. The AI buildout is producing vertically integrated compute-energy entities—companies that own the models, the data centers, the storage, and the generation under one roof. Three models of integration are forming: partnership, coalition, and full vertical control. Which one you build on determines which dependencies you carry.

In December, NextEra Energy and Google Cloud expanded their partnership to develop multiple gigawatt-scale data center campuses across the United States, each with co-located power generation purpose-built for the facility. NextEra already has roughly 3.5 GW in operation or contracted with Google. More interesting than the megawatts, though, is what’s happening to the utility itself. NextEra is taking Google’s AI models—TimesFM for time-series forecasting, WeatherNext for weather prediction—and applying them to its own grid operations through an internal initiative called Project REWIRE to run its power grid as a predictive system. Rather than simply adopting generic AI logic, NextEra is using these models (fine-tuned versions of them) on its own operational data and combining them with proprietary grid and optimization models. This allows the company to forecast demand, renewable output, equipment failures, and weather impacts in real time, improving reliability, reducing costs, and shifting from reactive operations to proactive decision-making—while also positioning this capability as a scalable product for other utilities.

The first commercial product is scheduled to launch on the Google Cloud Marketplace by mid-2026. Similar to how Amazon monetized its delivery network that was previously a cost center, NextEra is turning an internal operations system into a commercial software product, allowing it to sell predictive grid intelligence as a service. Petter Skantze, NextEra’s chief risk officer, oversees the program. A utility executive running an AI transformation that sells software on a hyperscaler’s platform: that is an organizational mutation rather than a procurement deal.

In parallel, at CERAWeek in March, NVIDIA and Emerald AI announced that six major US energy companies—AES, Constellation, Invenergy, NextEra, Nscale Energy & Power, and Vistra—would collaborate on flexible AI factories designed to operate as grid assets. NVIDIA’s Vera Rubin DSX reference design now includes DSX Flex, a software layer connecting compute workloads directly to power-grid services. Emerald’s Conductor platform demonstrated the ability to cut AI data center power consumption by 25% during grid stress events without degrading compute performance.

The math behind that 25% matters. Emerald estimates that if AI data centers can flex their power requirements by 25% for roughly 200 hours per year, that unlocks approximately 100 GW of stranded grid capacity across the US—more than the entire American nuclear fleet. The bottleneck, in other words, is not simply generation. It is also the inflexibility of the load. Utilities have been reluctant to approve large interconnections for AI facilities because a data center that draws 500 MW around the clock is a liability to grid stability. A data center that can throttle to 375 MW during peak stress, on software command, is something a utility can work with. That is why the orchestration software—the layer between compute and grid—has become the critical enabler for getting AI facilities connected faster.

Emerald was founded by Varun Sivaram, former CTO of ReNew Power in India and former senior US climate diplomat—a background that straddles exactly the energy-compute-sovereignty intersection. The coalition model accepts that compute and energy will remain under separate ownership and uses a shared software standard to coordinate between them.

It is worth noting what is taking shape here in aggregate.

NVIDIA, Emerald AI, Google, NextEra, and five other major US utilities are assembling a domestic compute-energy integration ecosystem where the software connecting AI workloads to grid services is becoming a proprietary American advantage. REWIRE, DSX Flex, and Emerald’s Conductor are not generic tools—they are trained on US grid data, validated against US interconnection rules, and sold through US cloud platforms. Countries building sovereign AI infrastructure today can import GPUs. They cannot easily import the orchestration layer that makes those GPUs grid-compatible at scale.

Then there is the Musk model.

When SpaceX acquired xAI in February for a combined $1.25 trillion, the headline was orbital data centers. The structure underneath is more telling. Across his companies, Musk now owns every layer of the AI infrastructure stack: compute (Colossus in Memphis, 200,000+ GPUs targeting one million), energy storage (Tesla Megapacks, over $375 million deployed at the Memphis site, with a 50 GWh/year Megafactory in Brookshire, Texas currently ramping up), power generation (a 1.2 GW dedicated plant, plus a joint venture with Solaris Energy carrying 1,140 MW in its orderbook), connectivity (Starlink, 9,000+ satellites), a proprietary data pipeline (X’s real-time feed), and end-user distribution (Grok embedded in Tesla vehicles).

Musk didn’t negotiate a PPA. He didn’t join an interconnection queue. He shipped a power plant across an ocean to Tennessee. That is vertical integration the technology industry hasn’t seen since AT&T owned the phones, the wires, and Bell Labs.

The three models sit on a spectrum of integration—and each carries different dependencies. In the partnership model, the orchestration layer (REWIRE, Emerald’s Conductor, NVIDIA’s DSX Flex) exists because the energy entity and the compute entity are separate organizations that need software to coordinate between them. In the coalition model, that software intermediary serves even more parties. In the vertically integrated model, the orchestration layer disappears entirely—there is no organizational boundary to mediate.

Build on the partnership model and you get speed, but you layer your hyperscaler’s AI deep into your energy operations. Build on the coalition model and you get vendor flexibility, but you depend on a reference architecture you don’t own. The fully integrated model eliminates those trade-offs—but it requires controlling rockets, batteries, generation, data, and distribution simultaneously, a portfolio that exists exactly once on earth.

For anyone making infrastructure or sovereignty decisions around AI, the organizational model matters as much as the technology specification. Signing a PPA with a utility is a different structural commitment than co-developing grid AI tools on a hyperscaler’s marketplace. And neither is the same as owning your own power plant. The question worth asking before your next capital allocation: does your current architecture create the optionality to deepen integration over time, or does it lock you into a permanent intermediary?

A note on independence: All opinions shared in this newsletter are my own and do not reflect the views of dmg events, ADIPEC, or any affiliated organizations. This is personal analysis, not institutional positioning.

Sources

• NextEra Energy / Google Cloud partnership press release, December 8, 2025

• NVIDIA and Emerald AI, CERAWeek announcement, March 23, 2026

• NVIDIA Vera Rubin DSX AI Factory reference design, GTC, March 16, 2026

• Emerald AI Conductor demonstration, EPRI DCFlex Initiative, Phoenix AZ (July 2025)

• Emerald AI $24.5M seed round (Radical Ventures, NVentures, CRV); $42.5M follow-on (October 2025)

• SpaceX–xAI merger, February 2, 2026 (CNBC, Bloomberg, TechCrunch)

• SemiAnalysis, “xAI’s Colossus 2,” September 2025

• NextEra Energy Q4 2025 earnings; January 2026 investor presentation; NEE leadership page (Skantze bio)

Governments around the world now treat compute as strategic infrastructure. The UAE’s Stargate campus is targeting 5 gigawatts of sovereign compute in Abu Dhabi, backed by G42, OpenAI, Oracle, and NVIDIA. Saudi Arabia’s HUMAIN is scaling one of the world’s largest government-owned facilities at Hexagon in Riyadh, a 480-megawatt Tier IV complex. India is scaling from 38,000 to 100,000 public GPUs by December 2026. France, Japan, Canada, and more than 20 other governments have signed sovereign AI partnerships with NVIDIA. But the structural question embedded in all of this capital is: what workload are they building for?

The answer, almost universally, is training — massive centralized clusters designed to produce AI’s foundation models. That assumption is about to collide with a workload shift that changes the entire calculation. Deloitte estimates that inference already accounts for two-thirds of all AI compute in 2026, up from one-third in 2023. McKinsey projects inference will represent over half of all AI compute and 30–40% of total data center demand by 2030. The inference chip market is projected to exceed $50 billion in 2026, a shift that prompted NVIDIA’s $20 billion 'reverse acquihire' of Groq’s inference unit, a move designed to monopolize the low-latency 'token-shuffling' layer that now defines the global AI stack.

Lenovo’s 2026 industry analysis projects that the 80/20 split between training and inference spending is reversing. Training a model is a one-time cost that does not directly generate revenue. Serving it to hundreds of millions of users via inference is a continuous, scaling cost that generates scaling revenue — and it is now the dominant workload in the global AI stack.

This matters because training and inference have fundamentally different infrastructure requirements — and different vulnerability profiles. Training workloads are geographically more flexible: build a massive cluster wherever power is cheap and baseload is guaranteed. Inference workloads follow users. They require low-latency, metro-adjacent infrastructure close to population centers. A piece published last week in Communications of the ACM argued that edge inference is increasingly a sovereignty problem, not a latency problem; regulatory mandates under the EU AI Act, India’s DPDP Act, and Brazil’s LGPD are pushing compute to the edge faster than latency requirements alone would dictate. The implication is significant: India’s 1.4 billion potential revenue-generating inference endpoints (read, people) begin to look more strategically consequential than a multi-gigawatt training campus. The country that hosts the users generates the sustained demand that funds the ecosystem. A country that hosts only the training cluster can technically repurpose it for inference — the same GPUs handle both workloads — but at significantly worse cost-per-token than dedicated inference silicon such as Groq’s LPU or Amazon’s Inferentia, in a facility optimized for remote cheap power rather than metro-adjacent low latency, and without the user base that generates continuous demand. The hardware is not stranded, but the strategic logic behind its location, economics, and demand source is.

The current conflict in the Gulf has made this asymmetry more visible. Any transparent assessment of a sovereign compute strategy requires examining at least three dimensions simultaneously: the political trust architecture (alignment with technology partners and in-country legal and regulatory clarity), the intelligence layer (access to models, data, and talent), and the physical power base (energy, capital, grid capacity, and physical security). The war, is stress-testing the power dimension in real time, so long as Hormuz remains functionally closed. The IEA warned this week that the disruption surpasses the combined oil crises of 1973 and 1979. For Gulf sovereign AI strategies funded by energy export revenue, the disruption cuts both directions: the revenue financing the buildout, and the physical infrastructure that sits in a conflict zone where three commercial data centers have already been hit. An inference-centric world distributes this risk differently. Distributed nodes across multiple geographies are harder to disable than a single mega-cluster — though they require a trust framework, energy supply, and governance architecture in every jurisdiction where they operate. The fragility shifts from concentration to connective tissue.

For infrastructure investors and policy planners, the practical takeaway is clear. The sovereign AI strategies announced over the past 18 months were designed for a training-centric model. The workload mix is shifting toward inference faster than those plans are adapting. Countries with large, digitally active populations and regulatory frameworks mandating local processing — India, Indonesia, Brazil — may find themselves in stronger structural positions than countries that optimized for raw training throughput. The European Data Centre Association’s 2026 report signals that future capacity growth will be constrained by grid readiness, not capital. Sovereign compute strategies that can flex between centralized training and distributed inference, while maintaining trust and security across both modes, will prove more resilient than those locked into a single architecture. The sovereignty question is no longer just “can you train a model?”, but is increasingly: where do the inferences run, who governs them, and what happens when the workload your infrastructure was built for is no longer the workload that matters?

A note on independence: All opinions shared in this newsletter are my own and do not reflect the views of dmg events, ADIPEC, or any affiliated organizations. This is personal analysis, not institutional positioning.

Sources

McKinsey, “The Future of AI Workloads,” February 2026

Deloitte, “Technology, Media and Telecom Predictions 2026: AI Compute”

Lenovo, CIO Playbook / 2026 Industry Analysis

JLL, “2026 Global Data Center Outlook”

European Data Centre Association, 2026 State of European Data Centers

ACM, “Inference at the Edge Is a Sovereignty Problem, Not a Latency Problem,” March 2026

G42/OpenAI/Oracle, Stargate UAE announcement, May 2025; The National, January 2026

Saudi Press Agency, Hexagon Data Center launch, January 2, 2026

India PIB, AI Impact Summit GPU announcements, February 2026

CNBC/NPR, Iran ceasefire developments, March 25, 2026

IEA Director Fatih Birol, remarks March 23, 2026

The instinct is to ask whether all of this slows the AI buildout. The critical question is which layers of the AI stack break under a sustained energy shock—and which accelerate because of it. On the software layer, the answer is unambiguous. The US military struck over 1,000 targets in the first 24 hours of Operation Epic Fury using Palantir’s Maven system integrated with AI for real-time targeting. Palantir stock is up 12% since the campaign began, its US government revenue having grown 55% year-over-year to $1.9 billion in 2025. In parallel, Bloomberg reported this week that Maxence Visseau, founder of investment firm Arkevium, used AI to cut research time by roughly 80%, stress-testing conflict scenarios across asset classes in minutes. When a CFO faces Oman crude above $165 and Brent above $110, compressing margins and spiking input costs, the business case for AI-driven efficiency stops being aspirational.

The hardware layer faces a different calculus. The Middle East AI buildout—one of the fastest-growing in the world through 2025—is obviously stalling. CSIS analysts now frame data centers as legitimate military targets in modern armed conflict. Pure Data Centre Group’s CEO told CNBC his firm would “slow down” in the region. The Atlantic Council’s Tess deBlanc-Knowles notes the Gulf remains attractive for its sovereign capital, energy, and Global South market access—but hyperscalers are running scenario analyses on redirecting next-wave capacity to Northern Europe, India, and Southeast Asia. This is geographic redistribution, though, not aggregate reduction. The CapEx still flows.

The stress-test matrix above maps where specific AI stack components begin to fracture, plotted against oil price and conflict duration. The thresholds are directional, anchored to observable data. Saudi Arabia’s fiscal breakeven sits at $78–94 per barrel depending on whether PIF spending is included (IMF, Bloomberg Economics), meaning Gulf sovereign wealth funds are currently generating windfall surpluses of $49–72 billion annualized. Goldman Sachs estimates European gas at €74/MWh would replicate 2022-level demand destruction. On March 18, TTF spiked to $24/MMBtu—approximately €75/MWh—touching Goldman’s crisis threshold intraday before settling lower, with Brent at $109. TTF has since traded in the €50–67/MWh range, trending toward the crisis line as South Pars damage compounds.

The direction matters more than any single close. Energy constitutes 15–25% of data center operating costs at baseline—but the South Pars escalation changes the timeline. Production infrastructure damage means LNG supply disruption persists beyond any ceasefire, repricing gas-dependent data center markets in Asia and Europe on a horizon measured in quarters, not weeks. Columbia’s Center on Global Energy Policy described March 18 as the moment the war moved “from logistical disruption to meaningful and potentially longer-lasting damage to core energy infrastructure assets.”

Duration remains the critical variable, but the South Pars strike has compressed the escalation curve. At three weeks with only shipping disruption, this was a market shock with geographic reshuffling. With production infrastructure now damaged, some portion of the LNG supply hit is locked in regardless of diplomatic outcomes. Extend the conflict to eight or sixteen weeks and the 1970s secular inflation analog—what Bloomberg’s Simon White calls “Act Three”—begins repricing every energy-linked asset class. Janus Henderson is already flagging that hyperscaler credit spreads, which started widening in Q3 2025, still don’t reflect the historical pattern where large debt-funded CapEx cycles produce meaningful spread widening.

The fertilizer disruption—roughly a third of global supply transits the Gulf—introduces a second-order food inflation channel that most AI infrastructure analyses ignore entirely. Several energy strategists, including Rystad and Financial Sense, frame the current disruption as the front edge of a structural energy supply cycle extending to 2030—which means this matrix is less a war-duration exercise than a preview of the operating environment for the next generation of AI infrastructure buildout.

For sovereign compute, the structural implication is clear. The gap between countries with installed AI infrastructure and those still building is widening in real time. The US benefits from its massive installed base and insulated domestic gas prices—Henry Hub near $3/MMBtu while Asian JKM trades above $20. India and Southeast Asia benefit as redirected buildout destinations. The Gulf states face a paradox: their sovereign capital has never been stronger, but the physical security case for hosting critical compute has weakened. Every government watching this conflict—from Brasília to Jakarta to Nairobi—is drawing the same conclusion: domestic compute capacity is a national security requirement, not a commercial preference. The war doesn’t slow the AI transition. It sorts who controls it.

Sources

• Columbia Center on Global Energy Policy, “Iran Conflict Brief: The High Cost of Attacking Energy Infrastructure,” March 20, 2026

• Gulf Mercantile Exchange (GME), Oman Crude OQD Marker Price, March 19, 2026 ($166.96/bbl)

• Fortune, “Oil prices hit nearly $110 as Iran vows to escalate the war,” March 18, 2026

• MacroVoices Episode #524, March 19, 2026

• EIA Short-Term Energy Outlook, March 10, 2026 (Brent $94/bbl as of March 9)

• CNBC, “How the Iran war could impact hyperscalers’ massive AI buildout in the Middle East,” March 11, 2026

• The Hill, “Iran war casts shadow over Middle East AI investments,” March 17, 2026

• Morgan Stanley, “AI Capex and the Iran War: Market Investing Risks,” March 2026

• Goldman Sachs, LNG supply and European gas price analysis, March 3, 2026

• Bloomberg, “Traders Overwhelmed by Iran News Are Turning to AI for Help,” March 19, 2026

• Janus Henderson, “How to play the AI mega-theme in fixed income,” March 2026

• IMF and Bloomberg Economics, Saudi Arabia fiscal breakeven estimates

• American Gas Association, Natural Gas Market Indicators, March 19, 2026

• CNBC, “Middle East war sends natural gas prices soaring,” March 3, 2026

• The Middle East Insider, “Saudi Arabia Economy March 2026,” March 14, 2026

• Rystad Energy, energy infrastructure supply chain analysis, March 2026

• Financial Sense, “Geopolitical Risk and Energy Infrastructure,” March 2026

A note on independence: All opinions shared in this newsletter are my own and do not reflect the views of dmg events, ADIPEC, or any affiliated organizations. This is personal analysis, not institutional positioning.

The data points are specific.

On March 1, Iranian drones struck three AWS data centers in the UAE and Bahrain — the first physical military attack on commercial cloud infrastructure. Two of three Availability Zones in AWS’s UAE region (me-central-1) went offline simultaneously, bypassing the multi-AZ redundancy that enterprises like Emirates NBD and Careem relied on for failover.

On March 3, the Pentagon issued a FASCSA (Federal Acquisition Supply Chain Security Act) designation against Anthropic — the first time this legal mechanism, historically reserved for foreign adversaries, was applied to an American company. Claude, the only AI model cleared for classified military networks, was ordered removed within 180 days.

As of mid-March, Claude is still running the Iran targeting cycle — processing roughly 1,000 potential targets per day, per CBS News — because there is no replacement ready.

Meanwhile, SpaceX handles over 90% of U.S. military satellite launches and its Starlink supports nearly 50 military commands. During the Trump-Musk dispute last summer, Musk floated decommissioning the Dragon capsule — America’s only crewed access to the ISS.

Three different infrastructure layers.

Three different vendors.

The same structural condition: single-provider concentration where the architecture assumed redundancy.

I’ve been developing an analytical framework I call the Trust–Intelligence–Power triangle, which scores sovereign compute strategies across three dimensions: Trust (political stability and alignment with technology partners), Intelligence (access to models, data, and talent), and Power (energy, capital, and physical security). The March events tested all three legs at once. The AWS strikes hit Power. The Anthropic phaseout hit Intelligence. And the dual-use problem — military AI running on commercial cloud, which gave Iran its stated targeting rationale — hit Trust. In each case, the vulnerability was concentration.

The Trust dimension deserves particular attention.

The IRGC said it struck AWS facilities because of their role in supporting military operations. Under the Geneva Conventions, civilian infrastructure that contributes to military action can become a lawful target. When the same cloud region hosts both consumer banking and classified targeting workflows, commercial SLAs effectively become military liabilities — a risk that neither the operator nor the customer originally priced.

Iran’s state media subsequently published a target list naming Amazon, Microsoft, Palantir, and Oracle.

Lloyd’s Joint War Committee has widened the Gulf high-risk zone, maritime war risk premiums have risen fivefold, and analysts project a 15–20% security premium on new Gulf data center capital expenditure for physical hardening alone. Reinsurers are debating whether cloud-based conflict falls under standard war exclusions or requires a new category entirely.

India is the near-term beneficiary.

Submarine cable infrastructure already connects Mumbai to the Gulf. IT load capacity stood at 1.4 GW as of mid-2025, and absorption rates in Mumbai and Chennai are pushing the 2026 forecast closer to 3.2 GW by year-end — ahead of schedule. AWS advised Gulf customers to migrate workloads to alternate regions, and the network path to India’s data center hubs already exists. But the architectural question goes further than geographic diversification. Every sovereign compute strategy built on a single foundation model, a single cloud region, and a single energy supply chain now faces a concrete version of a question that was previously theoretical: how does this architecture perform under pressure? If the world’s most capable military cannot swap AI vendors during an active conflict, the answer matters for Riyadh, New Delhi, and Brussels alike.

The sovereign compute strategies being funded today will operate for a decade or more. For infrastructure investors and policymakers, the question is practical: are the architectures being built today designed for the conditions that March 2026 is making visible?

A note on independence: All opinions shared in this newsletter are my own and do not reflect the views of dmg events, ADIPEC, or any affiliated organizations. This is personal analysis, not institutional positioning.

Sources

CBS News: Internal Pentagon memo orders military commanders to remove Anthropic AI technology from key systems

Mayer Brown: Anthropic Supply Chain Risk Designation Takes Effect

Euronews: Data centres are the new target in modern warfare during Iran war

Fortune: Iranian drone attacks on Amazon data centers signal a new kind of war

Just Security: Iranian Attacks on the Amazon Data Centers: A Legal Analysis

CNBC: Anthropic’s Claude would ‘pollute’ defense supply chain: Pentagon CTO

Lloyd’s List: Gulf war risk premiums topping double-digit millions of dollars per trip

The National Interest: US Space Strategy Can’t Rely on SpaceX Alone

CSIS: If Compute is the New Oil, War in the Gulf Significantly Raises the Stakes

Capacity: The Gulf gamble: Could the war in the Middle East drive a data centre exodus to India?

What happens when the Gulf deploys a form of geopolitical leverage it had not previously used: compute capital?

The consensus formed fast. Within days of Iranian drones striking three Amazon Web Services data centers across the UAE and Bahrain — the first confirmed military attack on hyperscale cloud infrastructure in any conflict — the narrative was set: the Gulf’s sovereign AI ambitions are physically fragile, and anyone building compute infrastructure in the region needs to rethink.

Bloomberg called the Strait of Hormuz “the hidden risk to the AI economy.” Fortune ran the drone strikes as a harbinger of future conflicts. CSIS drew parallels between oil and compute infrastructure as strategic targets. Rest of World mapped the simultaneous closure of both the Strait of Hormuz and the Red Sea to commercial shipping — the first time both data corridors between Asia and Europe have been shut at once. Semafor argued data centers are soft targets that need government-led protection.

The risks are real. But most of the coverage stops at the physical layer — and the more consequential shift is happening one layer down, in capital flows.

The question of whether Gulf data centers can withstand missile strikes is a valid engineering problem that will be addressed. The more consequential question is what happens when the Gulf deploys a form of geopolitical leverage it had not previously used: compute capital.

On March 5, the Financial Times reported that three of the four major Gulf economies — Saudi Arabia, the UAE, Kuwait, and Qatar — have begun internal reviews of overseas investment commitments. The review encompasses everything from existing contracts and force majeure clauses to future pledges made to foreign governments and corporations. A Gulf official told the FT the review was driven by converging pressures: declining energy revenues from disrupted exports, surging defense spending, collapsed tourism and aviation, and the sheer cost of intercepting hundreds of Iranian missiles and drones. An adviser to a Gulf government confirmed that the prospect of an investment pullback had already drawn attention inside the White House.

The scale of what is under review matters enormously for AI infrastructure. Gulf sovereign wealth funds collectively hold over $2 trillion in US assets. Following President Trump’s May 2025 regional visit, Saudi Arabia, the UAE, and Qatar pledged hundreds of billions of dollars in new US investments — commitments the White House promoted as validation of its Middle East posture. Those same pledges now sit in limbo. And a substantial share of them was earmarked specifically for AI infrastructure, data centers, and technology partnerships: MGX and BlackRock’s $30 billion AI infrastructure fund, PIF’s investments in SoftBank’s Vision Fund, KIA joining the AI Infrastructure Partnership, Mubadala’s stakes across robotics and compute, the planned Stargate UAE campus in Abu Dhabi, Amazon’s $5 billion Saudi AI hub.

This is structural capital in the US AI buildout, which became conditional.

Before the war started, Foreign Policy published a piece (February 23) arguing that Gulf AI investment in the United States was partly about securing American military protection. The logic was straightforward: sovereign wealth fund capital flowing into US technology created bilateral dependencies that made Washington more invested in Gulf security. It was an insurance premium paid in the form of infrastructure investment.

That deal broke its first real test. The Gulf states hosted US military bases, absorbed Iranian retaliation across their territory, watched drone strikes hit their ports, airports, refineries, hotels, and — for the first time in history — their data centers. The UAE alone intercepted 165 ballistic missiles, 2 cruise missiles, and 541 drones in a single weekend. QatarEnergy declared force majeure on LNG exports after strikes on the Ras Laffan facility, taking roughly 20% of global LNG supply offline. Saudi Arabia’s Ras Tanura refinery, processing 550,000 barrels per day, was hit by Iranian drones.

The investment review carries a political charge that extends well beyond budget management. QatarEnergy’s CEO Saad al-Kaabi was blunt in comments to the Financial Times: “Everybody that has not called for force majeure we expect will do so in the next few days that this continues. All exporters in the Gulf region will have to call force majeure. If this war continues for a few weeks, GDP growth around the world will be impacted.” The force majeure applies to delivery windows, not the long-term supply contracts themselves — a distinction that triggers different insurance clauses and changes the risk calculus for counterparties.

If Gulf sovereign wealth funds slow new commitments to US technology — or publicly invoke force majeure on existing contracts — the reverberations will reach every AI infrastructure deal structured around Gulf capital. The war exposed the Gulf’s dependency on US chips, which was already well understood. It simultaneously exposed the US’s dependency on Gulf capital, which was not. Hormuz made both visible at the same time.

The Physical Layer and the Economic Layer

The dominant narrative treats the Hormuz crisis as evidence that the Gulf’s sovereign AI model is broken. It isn’t. The model is stressed, and the connectivity layer is genuinely exposed. But the structural economics that made the Gulf attractive for AI infrastructure have not changed.

Electricity in Abu Dhabi still costs roughly $0.05 per kilowatt-hour, versus $0.09–$0.15 in Virginia. Both the UAE and Saudi Arabia have bypass pipelines for oil exports — the Abu Dhabi Crude Oil Pipeline to Fujairah, the Saudi East-West Pipeline — that can move part of the crude outside the Strait of Hormuz. Sovereign capital still does not require quarterly earnings calls. Permitting timelines measured in months, not years. These advantages do not disappear because of a conflict. They are why the Gulf attracted over $2 trillion in AI-related investment pledges in the first place.

Data centers are no more inherently vulnerable to ballistic missiles than oil refineries, desalination plants, or petrochemical complexes — all of which Gulf states have defended as critical national infrastructure for decades. The AWS drone strikes are an engineering and defense problem that will be solved the way Gulf states have solved similar problems for their energy infrastructure. Fortune reported that the US military used Anthropic’s Claude model, running on AWS infrastructure, for intelligence assessments and target identification during the Iran strikes. Iran’s Fars News Agency cited this dual-use reality as justification for targeting the Bahrain data center. The boundary between commercial cloud and military operations has collapsed — and that changes the threat model for every data center in a contested region. But it also changes the defense priority. Gulf states will now treat data center protection with the same urgency they apply to oil terminals and desalination plants.

The connectivity disruption from subsea cable exposure is a real vulnerability — but it is a connectivity vulnerability, separate from the compute capacity question. If anything, it validates the case for localized compute infrastructure rather than serving the Middle East and South Asia from data centers in Virginia. The organizations that bet on remote service delivery to one of the world’s fastest-growing AI markets are the ones most disrupted by the simultaneous closure of both maritime data corridors.

Who Is Actually Exposed

The Hormuz crisis reveals remote dependencies more starkly than local ones. Consider who has actually been hit hardest.

Bloomberg reported that more than half the world’s DRAM and NAND memory chips are manufactured in South Korea, and approximately 70% of advanced processing chips are made in Taiwan. Both countries are among the most dependent on Qatari LNG flowing through Hormuz. South Korea’s KOSPI crashed 12% in a single session — the largest drop since 2008 — with SK Hynix down 14.5%, hit harder than Samsung due to its higher reliance on specialized neon gas precursors that transit through the Strait. Taiwan holds less than a month of LNG reserves. Qatar alone accounts for roughly 30% of Taiwan’s LNG shipments. TSMC, which fabricates about 90% of the world’s cutting-edge semiconductors, consumes nearly 9% of Taiwan’s total electricity supply.

The implication is direct: the organizations that assumed they could serve the global AI buildout from fabs in East Asia powered by Gulf LNG are watching their energy supply chain fracture in real time. Taiwan’s gas reserves would last roughly ten to eleven days under normal consumption. Once ships currently en route discharge their cargoes, any ongoing disruption at Hormuz bites directly into the electricity supply powering the fabs that produce the world’s most advanced chips. The Thai stock exchange triggered its circuit breaker on an 8% drop. Asian jet fuel prices surged approximately 200%.

Europe’s exposure is no less severe. Natural gas prices nearly doubled within 48 hours of Hormuz closing. Germany’s wholesale electricity prices, already the highest among major economies, spiked further. The EU faces a reprise of its 2022 energy crisis, but this time with depleted winter gas storage. France (nuclear) and the Nordics (hydro) are partially insulated. The rest of the continent is not. Every European AI data center running on gas-fired electricity just got more expensive to operate.

The Hormuz crisis is a stress test of every jurisdiction’s energy assumptions. The jurisdictions that look most fragile are the ones furthest from the headlines.

Why Single-Layer Risk Analysis Fails

Most analysis of sovereign AI infrastructure collapses multiple, distinct risk layers into a single assessment. “Gulf AI: risky” is a headline, and headlines make poor investment frameworks.

Physical infrastructure resilience is one layer. Connectivity is another. Model and chip dependency is another. Regulatory sovereignty is another. Capital flows are yet another. The Hormuz crisis demonstrates why treating these as a single variable produces bad analysis. The Gulf’s energy and capital advantages remain structurally intact. Its connectivity layer is genuinely exposed. Its dependency on US chips and models was a known vulnerability before any missiles flew. And the bilateral trust relationship with Washington — the foundation of chip export approvals, the AI Acceleration Partnership, and the entire Gulf sovereign AI strategy — has been challenged, because the security assumption underlying the partnership failed its first real trial.

The Forward Signal

Whether this war lasts weeks or months, the compute capital dynamic is now visible and will not be forgotten. The Gulf states now understand that their AI infrastructure investment in the United States doubles as a geopolitical instrument. The US now understands that its AI infrastructure buildout carries a sovereign capital dependency it had not stress-tested. Neither side had priced this bilateral vulnerability before Hormuz. Both sides will price it from now on.

The practical consequences are already materializing. Every AI infrastructure partnership structured as “Gulf capital in exchange for US technology access” now carries a new risk premium. The chip export approvals that underpinned the Gulf’s AI strategy — the deal that let G42 divest Chinese holdings in exchange for NVIDIA access, the framework that gave HUMAIN its initial Blackwell allocation — are still technically in force. But the trust layer beneath them has been questioned. Regulatory approvals and commercial contracts assume stable bilateral relationships. When the bilateral relationship is under strain, every approval and every contract becomes contingent in ways that the original terms did not anticipate.

For anyone making deployment, investment, or regulatory decisions in the AI infrastructure space, a new question has joined the familiar ones about chips and electricity costs: who funds the stack — and what happens when that funding becomes conditional?

Sources

Financial Times (March 5, 2026): Gulf states reviewing investment commitments and force majeure clauses

Fortune (March 9, 2026): Iranian drone strikes on Amazon data centers — first deliberate targeting of data centers in conflict

Fortune (March 3, 2026): AWS confirms structural damage to UAE and Bahrain facilities

Bloomberg (March 5, 2026): “Hormuz Is the Hidden Risk to the AI Economy” — LNG dependency of chip fabs

Rest of World (March 5, 2026): Simultaneous closure of Hormuz and Red Sea; hyperscaler infrastructure at risk

Semafor (March 3, 2026): Data centers as soft targets requiring government-led protection

CSIS (March 3, 2026): “If Compute Is the New Oil” — parallels between oil and compute infrastructure

Foreign Policy (February 23, 2026): Gulf AI investment as partly about US military protection

CNBC (March 4, 2026): IRGC claims deliberate targeting of AWS Bahrain for military support role

CNBC (March 3, 2026): Banking and payments disruptions across UAE from AWS outages

QatarEnergy: Force majeure declaration on LNG shipments following Ras Laffan strikes

Khalaf Al Habtoor: Public criticism of Trump on X (March 5, 2026)

Tom’s Hardware (March 7, 2026): Uptime Institute confirms first military attack on hyperscale cloud provider

A note on independence: All opinions shared in this newsletter are my own and do not reflect the views of dmg events, ADIPEC, or any affiliated organizations. This is personal analysis, not institutional positioning.

The Anthropic-Pentagon confrontation last week demonstrated that the governance of AI models is a contested political space. Now extend that logic to agentic AI on critical infrastructure. The same model that a government might blacklist for political reasons could, in parallel, be managing grid stability for an allied country’s power system. The surface area for geopolitical disruption has expanded from data flows and chip exports to the operational control layer itself.

In 2024, more than fifty data centers in Northern Virginia tripped to backup generators simultaneously during a localized voltage instability event. PJM Interconnection’s automated grid controls — the Energy Management System and Automatic Generation Control — caught the cascading failure and rebalanced the system before it reached consumers. No blackout. No headlines. Human operators were aware of the alarm state, but the micro-second rebalancing decisions were made by the system before human intervention was possible. The operators acted on AI-synthesized data. The critical calls had already been made.

The entire debate about AI and energy has so far mainly focused on consumption — how many megawatts data centers need, where the power comes from, who pays for the grid upgrades. That’s a critical problem and a clear vulnerability, but it’s yesterday’s problem. The next one is harder: AI is beginning to fully operate the systems that produce, distribute, and manage energy. And the governance, liability, and insurance frameworks required for that shift do not yet exist.

ALREADY IN PRODUCTION

The deployments are further along than most realize. In December 2025, NextEra Energy — America’s largest utility — announced a landmark partnership with Google Cloud to deploy agentic AI across its field operations and grid management. The system integrates Google’s generative and agentic AI with NextEra’s asset data to predict equipment failures, autonomously schedule crew deployments around weather and supply chain disruptions, and optimize power flows using security-constrained modeling. The first commercial product launches on the Google Cloud Marketplace in the second half of 2026. Google’s open-source forecasting models — TimesFM 3.0 for time-series and WeatherNext 2 for weather prediction — will feed directly into grid operations. NextEra raised its earnings guidance on the strength of it.

In oil and gas, Baker Hughes launched Kantori in January 2026, a unified autonomous well construction system that uses AI and physics-based models to optimize drilling operations with “limited human intervention.” Its Cordant platform, now deployed with Aramco across four booster gas compression stations in Saudi Arabia, uses autonomous asset management to automate decision-making across asset health, process optimization, and energy management. These agents are currently bounded by physics-informed constraints — they cannot rewrite their own safety parameters — but the direction is clear: the industry is moving from tools that recommend to systems that act. Separately, PJM Interconnection partnered with Google’s Tapestry unit in April 2025 to rebuild its interconnection process using AI, deploying the “Twin Thread” model to unify dozens of databases into a single system that project developers, grid planners, and operators can query in real time. ERCOT, the Texas grid operator, created an entirely new Enterprise Data and AI Reliability Office in January 2026, explicitly designed to manage the intersection of autonomous systems and grid reliability.

And at Harvard, Professor Le Xie’s PowerAgent community — the first open-source initiative for agentic AI in power systems, featured as the cover story of IEEE Power and Energy Magazine — is building the foundational architecture: AI agents that can launch grid impact studies, run contingency analyses, evaluate interconnection requests, and generate operator reports, all through natural language interfaces that interact directly with engineering software via Anthropic’s Model Context Protocol. The key design feature is a “flexible human-in-the-loop”: the agent presents results for review rather than executing autonomously on critical decisions. But the direction of travel is unmistakable. The loop is getting shorter.

THREE WIDENING UNADDRESSED GAPS

The technology is moving faster than the institutions designed to govern it. Three structural gaps are opening, and each one represents both a risk for infrastructure investors and a strategic question for policymakers.

The first is the liability gap. When an autonomous AI agent makes a decision that affects grid stability — rerouting power flows, disconnecting a load, dispatching a crew — who is accountable? Today, there is no clear answer. The AI developer? The utility that deployed it? The cloud provider hosting the model? The grid operator that approved the protocol? Existing regulatory frameworks were built for a world where human operators make decisions and are accountable for them. FERC’s December 2025 order directing PJM to establish new rules for data center colocation didn’t address autonomous AI operations at all — because the agency is still focused on where power flows, not on who or what decides how it flows. The gap between operational reality and regulatory capacity is widening every quarter.

The second is the insurance gap — and this may be the most consequential bottleneck. Major insurers, led by AIG, have begun pulling back from AI liability coverage entirely. AIG’s AI exclusionary rider became standard for mid-market policies in late 2025. Lloyd’s of London syndicates are writing specialty AI policies through new ventures like Testudo and Armilla, but these products are expensive, narrow, and explicitly exclude AI developers and vendors. Lloyd’s began requiring algorithmic audits as a prerequisite for these policies in February 2026. The core problem, as Lexology documented in January 2026, is that insurers view AI systems as “black boxes” whose outputs are neither deterministic nor consistent — and AI errors propagate across user populations simultaneously, creating what underwriters classify as catastrophe exposure rather than standard liability. Model drift has been flagged as an uninsurable risk. A single model update that miscalculates grid loads across a utility’s entire territory isn’t an error. It’s a systemic event. For energy companies deploying agentic AI on critical infrastructure, the inability to insure autonomous decision-making is becoming a practical deployment constraint — not a theoretical concern.

The third is the feedback loop. Agentic AI running inference at the edge — in substations, on pipelines, at wellheads — changes where power demand materializes on the grid. It’s not just that AI consumes energy; it’s that AI’s own operational decisions reshape the energy system it’s managing. ERCOT’s new rules mandate that large loads above 75 MW participate in demand response programs and can be remotely disconnected during grid emergencies. But what happens when the AI agent managing a data center’s workload is also the AI agent optimizing the grid that feeds it? The optimization objectives may not align. The data center wants uptime. The grid wants stability. When both sides of that equation are being managed by autonomous systems operating faster than human review cycles, the conflict resolution mechanism doesn’t exist yet.

THE SOVEREIGN QUESTION

For countries building sovereign AI strategies, this adds a layer of dependency that the “Three Models” framework identified but didn’t fully explore. When NextEra deploys Google’s agentic AI to manage its grid, or when Aramco runs Baker Hughes’s Cordant on its gas infrastructure, the intelligence layer controlling physical operations is provided by a foreign vendor. The Gulf states’ technological dependence isn’t just about chips and foundation models anymore — it extends to the autonomous systems managing their most critical infrastructure. India’s application-layer strategy assumes foreign models are neutral tools. They are not. They are increasingly autonomous decision-makers with direct authority over physical systems. Whose safety protocols govern an AI agent controlling a grid switch in Riyadh? Whose liability framework applies? Whose insurance covers the failure?

The Anthropic-Pentagon confrontation demonstrated that the governance of AI models is a contested political space. Now extend that logic to agentic AI on critical infrastructure. The same model that a government might blacklist for political reasons could, in parallel, be managing grid stability for an allied country’s power system. The surface area for geopolitical disruption has expanded from data flows and chip exports to the operational control layer itself.

THE DECISION WINDOW

Within five years, no energy grid of significant scale will operate without autonomous AI agents. The technology trajectory is clear, the commercial incentives are overwhelming, and the deployments are already in production. The question is whether the governance, liability, and insurance frameworks will be in place before the systems are operating — or whether, as with every previous infrastructure revolution, the institutions will arrive a decade late.

For infrastructure investors, the insurance gap is the immediate signal. Any energy investment thesis that assumes agentic AI deployment at scale needs to model the insurance constraint — because if critical infrastructure can’t be underwritten, it can’t be financed. For policymakers, the regulatory gap is the urgent one: FERC, ERCOT, and PJM are redesigning interconnection rules for a world of AI-driven demand, but none of them have established frameworks for AI-driven operations. For energy executives, the workforce question is existential but deferred: the shift from “AI that recommends” to “AI that acts” redefines what operators do, and most organizations are not preparing their workforce for this transition.

The Virginia incident in 2024 ended well. The autonomous system caught the failure. But the next incident may involve an AI agent making a decision that a human operator would not have made — and when that happens, the question won’t be whether the technology works. It will be who answers for it. That question is unanswered today. And the systems are already running.

SOURCES

NextEra Energy / Google Cloud, “Landmark Strategic Energy and Technology Partnership,” Dec. 8, 2025

Baker Hughes, “Launch of Kantori™ Autonomous Well Construction Solution,” Jan. 28, 2026

Baker Hughes / Aramco, “Cordant™ APM Deployment Across Four BGCS,” Nov. 19, 2025

PJM / Google Tapestry, “Twin Thread” AI-driven interconnection partnership, Utility Dive, Apr. 10, 2025

ERCOT, “Strategic Organizational Changes,” Enterprise Data and AI Reliability Office, Dec. 12, 2025

FERC Order No. 2025-A, “Directs PJM to Create New Rules for Co-located Load,” Dec. 18, 2025

ERCOT / Texas SB6, large load interconnection rules (75 MW threshold), effective 2025

Zhang & Xie, “PowerAgent: A Road Map Toward Agentic Intelligence in Power Systems,” IEEE Power and Energy Magazine, vol. 23, no. 5, Sept.–Oct. 2025 (cover story)

Harvard SEAS, PowerAgent Community launch, PAI Symposium, May 2025

Kyndryl / PJM / Dominion Energy, “How AI is Reshaping Utilities and the Power Grid,” Feb. 2026 (Virginia incident)

Yale Clean Energy Forum, “Grid Modernization for Data Center and AI Loads,” Nov. 12, 2025

Lexology / Wiley, “When Insurance Won’t Cover AI,” Jan. 13, 2026 (model drift as uninsurable risk)

Metropolitan Risk Advisory, “Major Insurers Are Pulling Back from AI Liability,” Nov. 24, 2025 (AIG exclusionary rider)

Intelligent Insurer, “What Insurers Must Understand Before Underwriting the Next AI-Driven Catastrophe,” Feb. 3, 2026

Lloyd’s of London, algorithmic audit requirement for AI specialty policies, Feb. 2026

American Bar Association, “The Evolving Landscape of AI Insurance,” 2025

EDITORIAL NOTE

The Virginia data center incident is reported by Kyndryl citing PJM and Dominion Energy sources. PJM records indicate approximately 52 data centers were involved; the article uses “more than fifty” for accuracy. The micro-second rebalancing was supervised autonomy: operators were aware of the alarm state, but the system acted before human intervention was possible. The “three gaps” framework is the author’s analytical construct. Agentic AI deployments described are based on company announcements and press releases; actual operational scope and autonomy levels may differ from marketing descriptions. Baker Hughes’s Cordant agents are bounded by physics-informed constraints and cannot rewrite their own safety parameters. Insurance market analysis draws on legal and industry reporting; specific policy terms vary by carrier and jurisdiction. Gulf and India references draw on the author’s “Three Models of Sovereign AI” framework (Sovereign Compute, Feb. 2026).

This is a one-off analytical piece. It is not a recurring newsletter, not an investment advisory service, and not a solicitation. It presents a structural framework for understanding the AI infrastructure buildout and maps that framework to the sectors and asset classes where it is most directly expressed.

The tables below identify companies and assets that illustrate the thesis across different structural forces. They are not a portfolio, not a set of recommendations, and carry no suggested allocations or weightings. They are reference points for further research that you, the reader, can conduct — a starting map, not a set of directions.

The goal is to give you a way of thinking about AI infrastructure investing that you can apply on your own terms, with your own research, and in consultation with a licensed financial adviser.

The thesis: Five structural forces — energy demand, sovereign AI buildout, chip supply chain geopolitics, decentralised compute, and strategic mineral bottlenecks — are converging to reshape global capital flows. The companies and assets that sit at the intersection of these forces may define the next decade of returns.

1. The Macro Brief: The $650 Billion Collision

We are living through the largest infrastructure buildout in human history, and most investors are watching it through the wrong lens.

In February 2026, the four largest technology companies on earth collectively committed to spending approximately $625 billion in capital expenditures this year alone. Amazon will deploy $200 billion, Alphabet $185 billion, Meta $135 billion, and Microsoft $105 billion. These are not projections or aspirations. They are board-approved budgets already flowing into concrete, copper, and silicon.

This spending is colliding with a physical world that cannot absorb it. The US grid interconnection queue contains 2,600 GW of projects with a median wait of five years. The European Union has 1,700 GW stuck in similar delays. Seventy percent of American transmission lines are more than 25 years old. The April 2025 Iberian Peninsula blackout demonstrated what happens when aging grids meet exponential demand.

On February 25, NVIDIA reported its fiscal Q4 2026 results: $68.1 billion in revenue, up 73% year-over-year, with data center revenue hitting $62.3 billion. The company guided Q1 to $78 billion, beating estimates by $5.4 billion. The stock dropped 5.5% the following day, erasing $260 billion in market capitalisation. The market has shifted from asking whether AI infrastructure spending is real to asking whether the returns on that spending will materialise.

This is the central tension of 2026, and it creates the analytical framework this piece is built around.

The Five Forces in March 2026

Energy-AI Convergence. The IEA projects global data centre electricity consumption will double to 945 TWh by 2030, growing four times faster than total electricity demand. In the US alone, data centre power demand is projected to reach 134 GW by 2030, nearly tripling from today. This is not incremental growth; it is a step-change in how electricity is allocated globally.

Sovereign AI Infrastructure. Nations are treating compute capacity as a strategic asset. The UAE’s MGX fund targets $100 billion in AI and semiconductor investments. The G42/OpenAI/NVIDIA Stargate campus aims for 5 GW of capacity. Saudi Arabia’s HUMAIN subsidiary has secured $20 billion in partnerships targeting 6.6 GW by 2034. These are not venture bets; they are sovereign industrial policy backed by trillions in wealth fund capital.

Chip Supply Chain Geopolitics. The AI OVERWATCH Act advanced through committee in January 2026, proposing a statutory two-year ban on Blackwell-class chip exports to China. This is a material escalation from administrative BIS controls. Meanwhile, the hyperscaler shift toward custom ASICs is fragmenting what was a near-monopoly: the AMD-Meta five-year deal, valued at $60–$100 billion, signals the beginning of the inference era where custom silicon competes with merchant GPUs.

Decentralised Compute. Decentralised GPU networks are generating real revenue. The DePIN sector produced approximately $150 million in on-chain revenue in January 2026. Render Network has expanded beyond rendering into AI inference through its Dispersed subnet. Akash Network’s Burn-Mint Equilibrium code reached completion in February 2026 with mainnet launch scheduled for March 30. These are no longer speculative protocols; they are nascent utility markets. But they remain small relative to hyperscaler revenues, and intellectual honesty demands we acknowledge that.

Strategic Minerals. The top three refining nations control 86% of key energy minerals. China holds 80%+ of global graphite and rare earth processing. The IEA projects copper and lithium deficits of 30–40% by 2035. Every GPU, every data centre, every grid upgrade depends on physical materials whose supply chains are geographically concentrated and politically vulnerable.

What the Market Is Missing

Most AI investment analysis focuses on software companies and chip designers. The real bottleneck — and therefore the real analytical gap — is in the physical infrastructure layer: the power companies signing data centre contracts, the nuclear operators restarting reactors, the copper miners struggling to meet demand, the networking companies building the backend fabric that connects millions of GPUs.

This piece maps the full stack. From uranium in the ground to tokens on-chain.

2. Thesis Deep Dive: The Inference Rotation

Why the AI Hardware Winners of 2025 May Not Be the Winners of 2027

The DeepSeek R1 release in early 2025 proved that frontier-level AI performance could be achieved at one-tenth the compute cost of comparable models. This single development reshaped the investment landscape for AI infrastructure in ways the market has not yet fully absorbed.

The immediate reaction was a $593 billion wipeout in NVIDIA’s market capitalisation. But the structural implication runs deeper than one company’s stock price. DeepSeek demonstrated that the AI compute market is bifurcating: training workloads remain concentrated among a handful of frontier labs willing to spend billions on GPU clusters, while inference workloads — running models in production for billions of users — are shifting toward efficiency, custom silicon, and cost optimisation.

This is not a bearish development for AI infrastructure spending. The Jevons Paradox applies: when compute becomes cheaper per unit, total demand expands because more applications become economically viable. But it does change who captures value.

The Training-to-Inference Shift

NVIDIA built its dominance on training. Its CUDA software ecosystem and GPU architectures are purpose-built for the massively parallel matrix operations that training requires. For training, NVIDIA remains unchallenged. No serious competitor has emerged for frontier model training at scale.

Inference is different. Running a trained model in production requires different optimisations: lower latency, higher throughput per watt, lower cost per token. These requirements favour custom Application-Specific Integrated Circuits designed for specific workloads rather than general-purpose GPUs designed for everything.

This is where Broadcom has quietly become the most important company in AI that most investors have never heard of. Broadcom designs the custom ASICs that hyperscalers use for inference: Google’s Tensor Processing Units, Meta’s custom accelerators, and most recently OpenAI’s “Titan” chips under a partnership estimated at $100–200 billion over multiple years. Broadcom’s AI revenue is projected to hit $40 billion in fiscal 2026, with a $73 billion hardware backlog providing multi-year visibility. The February 2026 AMD-Meta five-year deal, valued at $60–100 billion, further confirms that hyperscalers are diversifying away from NVIDIA’s merchant GPU model.

The inference shift also creates opportunity in the decentralised compute layer. Networks like Render and Akash are fundamentally inference platforms. They cannot compete with NVIDIA’s DGX clusters for frontier training. But for running models in production at low cost, with geographic distribution and without hyperscaler lock-in, they offer a structural alternative that becomes more compelling as inference demand explodes.

The Analytical Implication

The inference rotation is not a threat to AI infrastructure spending. It is a redistribution of who captures value within that spending. Understanding this redistribution — and identifying which companies and assets are positioned for it — requires mapping the full stack: from the chip designers and contract manufacturers building the custom silicon, to the power companies and data centre operators housing it, to the decentralised networks offering an alternative distribution layer.

3. The Full-Stack Framework: Mapping Forces to Sectors

The core design principle behind this framework is that an investment thesis should be directly expressed through the assets it identifies. When you read about grid constraints, the relevant sector is power generation. When you read about export controls, the relevant sector is semiconductor fabrication. When you read about mineral bottlenecks, the relevant sector is upstream mining. When the thesis and the research point in the same direction, the analytical alignment is strong.

The tables below organise the five structural forces into three categories — infrastructure backbone, thematic plays, and digital asset infrastructure — and identify the companies and assets most directly connected to each force. These are analytical reference points for further research, not recommendations or a portfolio. No weightings or allocations are implied. Readers should conduct their own due diligence and consult a licensed financial adviser.

Category 1: Infrastructure Backbone

These names directly capture the physical forces the thesis describes. They rise and fall based on actual infrastructure buildout, not sentiment. This is where thesis-to-sector alignment is strongest.

Category 2: Thematic Plays

These names connect to themes where domain-specific intelligence — from energy conferences, sovereign wealth fund activity, and ASEAN energy policy — provides analytical context that generic equity research may not capture.

Category 3: Digital Asset Infrastructure

The Sovereign Compute thesis extends to the decentralised compute and on-chain infrastructure layer. These assets represent the digital side of the same structural forces — distributed GPU compute, cross-chain settlement infrastructure, and decentralised AI model training. Digital assets are highly volatile, speculative, and may result in total loss of capital. The names below illustrate where the thesis intersects with the digital asset ecosystem; they are not recommendations to purchase or hold any asset.

Force-to-Sector Alignment

The table below maps each structural force to the sectors and asset categories where it is most directly expressed. This is the core of the analytical framework — the idea that the investment thesis and the assets you research should point in the same direction.

4. Risk Framework

Intellectual honesty demands that any analytical framework acknowledge what could make it wrong. The following risks could undermine or invalidate the thesis above.

The bear case is real. If AI capex slows materially and crypto enters a prolonged downturn simultaneously, assets connected to this thesis would face significant drawdowns. The infrastructure-heavy sectors (power, semis, data centres) would fare better than digital assets in that scenario, but neither would be immune. Any framework that tells you only the bull case is not worth your time.

5. Catalyst Calendar

The following events represent known structural catalysts that could materially affect the thesis over the coming months.

6. How to Apply This Framework

This piece gives you a lens. What you do with it is up to you.

The five structural forces — energy demand, chip supply chains, data centre infrastructure, strategic minerals, and sovereign AI buildout — are the analytical backbone. When you read news about any of these forces, ask: which sectors and assets are most directly connected? The force-to-sector alignment table above is your map.

If the thesis resonates, your next steps are your own research. Examine the companies and assets named above. Read their earnings transcripts. Assess whether the structural force behind each name is strengthening or weakening. Decide which sectors match your own risk tolerance, time horizon, and financial situation. Consult a licensed financial adviser before making any investment decisions.

The framework is the contribution. The decisions are yours.

Disclaimer & Disclosures

This is a one-off analytical piece published for general educational and informational purposes only. It is not a recurring advisory service, not a managed portfolio, and not a solicitation to buy, sell, or hold any asset. Nothing in this publication constitutes personalised investment advice, financial advice, virtual asset advisory services, or a personal recommendation to any reader.

The companies, ETFs, and digital assets named in the tables above are identified as analytical reference points to illustrate the structural thesis. They are not recommendations. No allocations, weightings, or position sizes are suggested or implied. Readers are solely responsible for their own investment decisions and should conduct independent research and consult a licensed financial adviser before making any investment.

Cryptocurrency and digital asset investments are highly volatile and speculative. The value of such investments may decline significantly and may result in total loss of capital. Past performance of any asset discussed does not guarantee, predict, or indicate future results.

This piece is published by Ivan Ferrari in a personal capacity. Ivan Ferrari is VP of ADIPEC & Business Development at dmg events. This publication operates independently of dmg events and does not represent the views or analysis of dmg events or any affiliated entity. Ivan Ferrari may hold personal positions in assets discussed in this piece.

This publication is not regulated by VARA, the SCA, or any financial regulatory authority. It does not provide suitability assessments, maintain client agreements, or offer tailored recommendations to any reader.

SOVEREIGN COMPUTE

sovereigncompute.substack.com

© 2026 Ivan Ferrari. All rights reserved.

“The Gulf’s response to Iran’s retaliatory strikes demonstrated exactly why the region remains a serious contender for global AI infrastructure. ”

In the span of three weeks this February, the Pentagon airlifted a nuclear reactor to Utah on a C-17, a British consortium pitched floating nuclear power stations to be docked at US naval bases, and SpaceX filed with the FCC for an expanded constellation of up to 30,000 Gen2 nodes capable of dedicated edge-compute and orbital data processing. Meanwhile, in Malaysia’s Johor state, Chinese engineers were training AI models on rented Nvidia servers using physically transported hard drives, in data centers that don’t classify those servers as controlled items. These aren’t unrelated events. They are five expressions of a single structural logic: AI infrastructure is migrating to wherever the rules are thinnest — on land, at sea, in orbit, and across borders.

The pattern is regulation arbitrage — the systematic relocation of energy-intensive compute to jurisdictions where existing oversight either doesn’t apply or can be bypassed. In the United States, it runs along a spectrum. Behind-the-meter gas generation lets data centers skip interconnection queues and avoid utility commissions. Cleanview, an energy analytics firm, has identified 47 such projects representing approximately 50 GW of capacity, 90% announced in 2025 alone. McKinsey estimates 25–33% of all incremental data center demand through 2030 will be met behind the meter. One project in Texas will consume more power than all of Chicago without connecting to a single transmission line.

Further along the spectrum, military-jurisdictional nuclear bypasses the NRC entirely. The Valar Atomics airlift on February 15 — three C-17s carrying the Ward250 reactor to Utah, with Energy Secretary Chris Wright aboard — was explicitly framed as deregulation in action. The DOE’s Reactor Pilot Program routes licensing through the Department of Energy rather than the NRC, which can take a decade. Core Power, a London-based nuclear startup, is separately negotiating to dock floating reactors at military installations, drawing on naval designs that bypass civilian oversight. At the far end, SpaceX’s orbital filing proposes to exit the terrestrial regulatory stack entirely — no zoning, no water rights, no air quality review.

The Trump administration has made this logic explicit. David Sacks told a Davos audience in January that the president’s vision is to “let the AI companies become power companies.” The December 2025 executive order preempts state AI regulation; more than 300 state data center bills have been filed in the first six weeks of 2026, according to MultiState, but the federal government is simultaneously encouraging the industry to bypass state authority and generate its own power off-grid.

But the arbitrage isn’t only domestic. Southeast Asia is the international version of the same logic. Across six ASEAN economies, American and Chinese hyperscalers have committed over $50 billion in data center infrastructure, building side by side on the same grids under the same loose rules. Only one of ten ASEAN member states — Vietnam, as of December 2025 — has enacted the Digital Technology Industry Law, which mandates binding security standards for AI developers. The rest rely on non-binding governance guidelines with no enforcement mechanism. For hyperscalers, this means speed. For geopolitical strategists, it means something else: a theater where US export controls are structurally difficult to enforce.

But regulation exists for reasons that don’t disappear when you move the infrastructure. The xAI precedent in Memphis illustrates what happens when speed overrides oversight. Musk’s AI company built the Colossus data center in 122 days in 2024 by deploying up to 35 unpermitted methane gas turbines in a predominantly Black neighborhood that the American Lung Association had already rated “F” for ozone pollution. The EPA ruled in January 2026 that xAI violated the Clean Air Act. The NAACP filed suit. This is the template that other developers are now replicating — build the power plant on site, stay off the grid, minimize regulatory contact. The Washington Post reported on February 19 that dozens of off-grid projects planned across Texas, New Mexico, Pennsylvania, Wyoming, and Tennessee follow the same playbook.

This dynamic has no real parallel elsewhere — and that’s the point. Each major compute geography carries a structural vulnerability that defines how its AI infrastructure gets built, and each resolves the tension differently. The Gulf states — Saudi Arabia, the UAE, Qatar — don’t face the regulatory friction that drives American regulation arbitrage. State-led capital and sovereign energy eliminate community-level opposition by design. The Gulf’s structural advantage is speed, capital depth, and an accommodating regulatory environment. Its structural vulnerability is geographic: proximity to a neighbor with ballistic missile capability and a declared willingness to use it. That vulnerability moved from theoretical to concrete on February 28, when Iranian retaliatory strikes sent 137 missiles and 209 drones toward the UAE alone, shutting Dubai and Abu Dhabi airports, closing airspace across the Gulf, and triggering Strait of Hormuz closure warnings — the same corridor adjacent to submarine cable clusters carrying an estimated 17% of global internet traffic.

The Gulf’s response demonstrated exactly why the region remains a serious contender for global AI infrastructure. The UAE Ministry of Defense confirmed that air defenses intercepted the vast majority of incoming threats with high efficiency and reported no significant material damage to critical infrastructure. Cybersecurity systems remained operational around the clock. Supply chains held. Essential services continued. The UAE’s integrated emergency management system activated rapid-response protocols and maintained public order throughout. These are not the hallmarks of fragile states. They are the hallmarks of states that have invested heavily in both digital and physical security architectures — and the reason that commitments like Stargate UAE’s 5GW campus, Microsoft’s $15.2 billion investment, and KKR’s $5 billion data center partnership were made in the first place. As CSIS argued one day before the strikes, if compute truly is the new oil, the Gulf’s growing AI role will bind Washington to the region more tightly than current doctrine acknowledges.

India resolves the tension differently still. Its AI buildout operates within a permitting environment explicitly designed to attract data center investment — in February 2026, India announced a 20-year tax break for domestically built data centers. India’s vulnerability is not regulatory friction or geographic exposure but dependency: the world’s largest AI deployment infrastructure built entirely on foreign models and foreign chips, with Indian languages comprising less than 1% of global training data. The European Union is moving in the opposite direction, with the Energy Efficiency Directive requiring data center operators to report power and water usage, noncompliant facilities facing fines and restricted grid access.

The result, globally, is not a single system but a fracturing of the compute landscape into concentric rings of oversight. From fully regulated grid-connected facilities in Northern Virginia, to behind-the-meter gas plants operating in regulatory grey zones across Texas and Appalachia, to military-jurisdictional nuclear at sea, to the Gulf’s sovereign-capital model now stress-tested by kinetic conflict, to orbital infrastructure governed by spectrum licenses and little else. Each ring further from the center trades one form of accountability for another form of speed.

For infrastructure investors, energy executives, and policy strategists, the implication cuts both ways. The US model is most vulnerable to its own citizens at the ballot box. The Gulf model is most vulnerable to its geographic neighborhood — though this weekend demonstrated it has the defensive architecture to absorb the shock. The geography of compute is now shaped less by where the power is than by where the rules, the capital, and the security architectures align. The AI industry’s regulatory strategy is becoming, in effect: if you can’t change the rules, change the jurisdiction. The question is what gets lost in transit.

Editorial note: Gulf, Indian, and EU comparisons draw on the author’s “Three Models of Sovereign AI” framework (Sovereign Compute, Feb. 2026). The situation in Iran and the Gulf remains fluid; this analysis focuses on structural infrastructure implications, not the conflict itself. Orbital data center timelines remain speculative; the analysis focuses on what these proposals reveal about terrestrial constraints.

Sources: Washington Post (Feb. 19, 2026); Cleanview/Distilled Earth (Feb. 2026); AP/PBS/NPR (Feb. 21, 2026); GeekWire (Jan. 31, 2026); Maritime Executive (Feb. 23, 2026); CNBC (Jan. 16, 2026); Wall Street Journal (Jun. 2025); McKinsey/EDB “AI in Southeast Asia” (Feb. 2026); CSIS “Beyond the Matrix” (2025); MultiState (Feb. 20, 2026); CreditSights (Feb. 2026); Bain & Co Malaysia data center analysis; Stars and Stripes (Jan. 2, 2026).

A note on independence: All opinions shared in this newsletter are my own and do not reflect the views of dmg events, ADIPEC, or any affiliated organizations. This is personal analysis, not institutional positioning.

The US controls its domestic stack. China is building a parallel one. The Gulf buys into the American ecosystem under strict export conditions. India bets on applications over models.

But Southeast Asia — 700 million people, six fast-growing digital economies, and a region where binding AI regulation has only just begun to emerge in Vietnam, leaving the rest as a high-speed, non-binding laboratory — is the one place where American and Chinese hyperscalers are building data centers side by side, competing for the same customers, on the same grids, under the same loose rules.

This is not a metaphor.

AWS, Microsoft, and Google have collectively committed over $50 billion to the region. Alibaba Cloud, Tencent Cloud, and ByteDance are building alongside them — ByteDance alone has earmarked $2.1 billion for an AI hub in Malaysia’s Johor state, where it is anchor tenant at Bridge Data Centres’ 110MW hyperscale facility. GDS International, a Singapore-listed subsidiary of a Chinese parent, reports that roughly 70% of its Southeast Asian client mix is Chinese.

These figures still look modest against the $600-plus billion in projected 2026 hyperscaler capex globally — but that comparison is misleading. Roughly half of US hyperscaler spend is a check to Nvidia and TSMC for silicon; it never touches local soil. The billions landing in Southeast Asia are steel, concrete, and power connections — permanent infrastructure with a local GDP impact orders of magnitude higher per dollar than a GPU purchase order routed through Santa Clara.

The region is absorbing both stacks simultaneously, and no one is forcing a choice. Indonesia’s Tokopedia runs its live video and data analytics on Google Cloud while its core transactional infrastructure sits on Alibaba Cloud data centers in Jakarta, and in 2024, TikTok reentered the Indonesian market through a merger with Tokopedia itself. One company, both cloud stacks, both superpowers’ consumer platforms, no regulatory friction.

The structural reason is regulatory.

ASEAN’s Guide on AI Governance and Ethics, updated in January 2025 to cover generative AI, is non-binding. No Southeast Asian country had a binding AI law until Vietnam passed one in December 2025. Malaysia, Thailand, and Indonesia are all drafting frameworks, but none have enacted hard regulations. For hyperscalers, that translates to speed: faster permitting, fewer compliance layers, no equivalent of the EU AI Act’s risk classifications. For geopolitical strategists, it translates to something else — a theater where export controls are structurally difficult to enforce.

The evidence is already here. The Wall Street Journal reported in June 2025 that four Chinese engineers flew from Beijing to Kuala Lumpur carrying suitcases packed with 60 hard drives — 80 terabytes of AI training data — and trained models on 300 rented Nvidia servers at a Malaysian data center.

Malaysia does not classify Nvidia-powered servers as controlled items under its Strategic Trade Act. CSIS noted that the performance gap between the best US and Chinese AI models shrank from 9.3% in 2024 to 1.7% by early 2025. Southeast Asia’s under-regulated data centers may be accelerating that convergence — particularly after DeepSeek-R1’s January 2025 release demonstrated that frontier-competitive models could be trained on far fewer GPUs, making the suitcase-and-rented-server approach more viable than ever.

The energy math compounds the stakes. Malaysia’s data center power demand is projected to hit 21% of national electricity capacity by 2027, a massive leap from 7% in 2022, according to Bain & Co. By 2035, the government warns this could exceed 50% of the Peninsular grid’s total capacity. Johor state alone has 5,800MW of planned capacity — enough to power nearly 5 million households. The national grid is 92% fossil fuel. Malaysia has imposed new power tariffs on large data centers, and water shortages in Johor and Selangor have already forced authorities to slow approvals.

For Pax Silica signatories — Singapore, the UAE, and now India are all members — the implications are direct. Singapore has announced tranches aimed at expanding beyond its current 1.4GW data center base, pushing larger requirements to neighboring Malaysia. The Johor-Singapore Special Economic Zone creates passport-free travel between the two, effectively extending Singapore’s digital economy into Malaysian power grids.

Southeast Asia is not a sideshow in the sovereign compute race. It is the contested middle ground — the one place where both superpowers are building, both regulatory models coexist, and the energy constraints haven’t yet forced a reckoning. Infrastructure investors reading Pax Silica as a neat division of the world into blocs should look more carefully at Johor. The blocs overlap there. And the grid is already groaning.

Sources: McKinsey/EDB “AI in Southeast Asia” (Feb 2026); Bain/Google/Temasek “e-Conomy SEA” (2025); CSIS “Cloud Computing in Southeast Asia and Digital Competition with China” (Oct 2024); CSIS “Beyond the Matrix: AI Governance Gaps in Southeast Asia” (2025); Wall Street Journal (Jun 2025); CNBC Johor data center investigation (Aug 2025); Asia Society Policy Institute (Jan 2026); CBRE Asia Pacific Data Centre Outlook (2026); ISEAS-Yusof Ishak Institute (Feb 2026); Crowell & Moring ASEAN Digital Ministers analysis (2026).

A note on independence: All opinions shared in this newsletter are my own and do not reflect the views of dmg events, ADIPEC, or any affiliated organizations. This is personal analysis, not institutional positioning.

China is conspicuously absent from this analysis — not because it doesn’t matter, but because it’s playing a different game entirely. Beijing is building a parallel AI stack outside the US-controlled supply chain, forced by export controls into a self-sufficiency model that deserves its own treatment. This piece focuses on the three major models competing within the US-anchored ecosystem — because that’s where the infrastructure capital, the energy bottlenecks, and the policy contradictions are most acute right now.

The American Model: Control the Chokepoints

The US doesn’t build sovereign AI infrastructure the way other countries do. It doesn’t need to. American sovereignty comes from controlling the supply chain itself: the chips, the cloud platforms, the foundation models, and the export licenses that determine who else gets access. NVIDIA, Microsoft, Google, Amazon, and Meta are projected to spend on the order of $650–700 billion in capex in 2026, with the majority tied to AI infrastructure. Hyperscaler earnings make clear the market is supply-constrained, not demand-constrained. Microsoft disclosed an approximately $80 billion backlog of Azure orders it cannot fulfill because there isn’t enough power.

Sources: Introl hyperscaler capex analysis (Jan 2026); MacroNotes analysis of Microsoft Azure backlog.

The Trump administration has turned this into explicit policy. Through executive action in January 2025 — revoking Biden’s AI safety executive order (EO 14110) and replacing it with “Removing Barriers to American Leadership in Artificial Intelligence” (EO 14179) — and a sweeping December 2025 order (“Ensuring a National Policy Framework for Artificial Intelligence”) preempting state-level AI regulation, the administration has framed sovereignty as deregulation. The December order directs the Attorney General to establish an AI Litigation Task Force within 30 days to challenge state AI laws deemed inconsistent with federal policy. It instructs the Secretary of Commerce to condition remaining broadband deployment (BEAD) funds on states not having “onerous” AI laws. The message is clear: the federal government will clear the runway for private capital to build at scale, and use export controls to decide who else gets to play.

Sources: Mayer Brown analysis (Dec 23, 2025); Squire Patton Boggs analysis (Jan 2025); Utility Dive (Dec 8, 2025).

The blind spot: the people who live next to the runways are revolting

A populist backlash is brewing across red and blue states alike. Senator Bernie Sanders endorsed a national moratorium on data center construction in December 2025, likely the first member of Congress to do so, citing concerns about the unregulated growth of AI and its impact on employment. Governor Ron DeSantis announced a proposed AI Bill of Rights on December 4, 2025, including provisions that would prohibit utilities from charging Florida residents more to support hyperscale data center development and allow local governments to block data center construction. In Wisconsin, the mayor of Port Washington faced a recall campaign after approving a $15 billion data center for OpenAI and Oracle through the Stargate project. In the PJM Interconnection — the grid serving 67 million people across 13 states — data center demand added over $23 billion to capacity auction costs across three consecutive auctions, costs that flow directly to consumers. According to Data Center Watch, $64 billion in data center projects were blocked or delayed through early 2025 by local bipartisan opposition. Residential electricity prices rose approximately 5% in 2025 and are forecast to climb another 4% in 2026, according to the federal Energy Information Administration. With mid-term elections in November, data center opposition is becoming a bipartisan campaign issue.

Sources: E&E News/Politico (Dec 17, 2025); Florida Governor press release (Dec 4, 2025); Wisconsin Watch/Racine County Eye (Jan 2026); Bloomberg/Monitoring Analytics (Jan 5, 2026); CNBC (Jan 1, 2026); EIA electricity price forecast.

The paradox is sharp: the administration that removed every federal barrier to AI buildout may find the real barrier is voters.

The Gulf Model: Convert Energy Wealth into Compute Wealth

Saudi Arabia and the UAE are running a different playbook. Where the US controls the supply chain, the Gulf is investing its way into the infrastructure layer — fast, at scale, and with sovereign capital that doesn’t need quarterly earnings calls.

The headline commitments are staggering. G42 and partners have announced plans for a multi-gigawatt UAE-US AI Campus in Abu Dhabi, with an initial 1GW phase breaking ground. Microsoft committed billions to the UAE through 2029. Saudi Arabia’s HUMAIN has announced plans to deploy substantial compute capacity by 2030, beginning with an initial allocation of Nvidia Blackwell chips. At a business event in 2025, Saudi Arabia attracted over $20 billion in headline investment commitments, including Groq’s announced plans for a major inference data center.

Note: Gulf investment figures represent headline commitments and announced plans; finalized deal terms may differ.

The structural advantage is deceptively simple: typical Gulf electricity costs run approximately $0.05–0.06 per kilowatt-hour versus $0.09–0.15 in the US. At data center scale, that 50–70% energy cost advantage compounds into an operating margin moat. The Gulf is essentially doing what it did with petrochemicals in the 1980s: converting a natural resource advantage (cheap energy) into downstream industrial capacity (compute), backed by state capital.

Source: Euro-Security analysis of AI infrastructure economics (2026).

Governments don’t just enable this buildout — they orchestrate it. The UAE was the first country to appoint a Minister of State for Artificial Intelligence, in 2017. Abu Dhabi established the Artificial Intelligence and Advanced Technology Council. Saudi Arabia’s SDAIA coordinates national AI strategy under Vision 2030. Neither country has a standalone AI law, but both govern through a layered approach — data protection, ethics charters, cybersecurity rules, and sector-specific guidance — that gives the state flexibility to adjust without locking in rigid frameworks. The UAE’s approach has been described as a “sandbox state”: policy precedes statute, enabling rapid iteration.

The blind spot: technological dependence

Every advanced chip, every foundation model, every cloud operating system comes from the US or (to a lesser extent) China. When the Trump administration approved advanced AI chip exports to the UAE and Saudi Arabia, it attached strict security requirements and reporting obligations. G42 divested its Chinese investments — including a reported stake in ByteDance — under US national security pressure. The Gulf’s AI sovereignty is real at the infrastructure layer but fragile at the intelligence layer. If export policy shifts, or if US-China competition forces harder choices about alignment, the Gulf’s compute campuses could face the same bottleneck that constrains their clients: access to the models and chips that make the hardware useful.

Sources: Multiple reporting on G42 divestment from Chinese investments under US pressure (2024–2025).

Citizens in the Gulf, meanwhile, aren’t pushing back — partly because energy is heavily state-subsidized and electricity costs are largely invisible to households, and partly because the state-led model doesn’t create the community-level disruptions that American data centers do. There are no recall campaigns in Abu Dhabi.

The Indian Model: Own the Application Layer

India’s play is the most counterintuitive. While the US controls the supply chain and the Gulf builds the hardware, India is betting that owning the application layer — deploying AI at population scale — matters more than owning the models or the chips.

The India AI Impact Summit made the strategy explicit. Adani Group announced plans for a massive, multi-decade investment in renewable-powered AI data centers. The Indian government expects over $200 billion in AI investment across the full stack in the coming years. Google, Microsoft, Amazon, OpenAI, and Anthropic all sent their CEOs to New Delhi. India has earmarked roughly $1 billion for an AI-focused venture capital fund. Micron is establishing semiconductor packaging and test operations in Gujarat.

Note: Indian investment figures represent government projections and corporate announcements; actual deployment timelines may vary.

But India has consciously chosen not to chase frontier model development. The country’s Economic Survey has urged a focus on application-led innovation rather than competing on foundation models. As Infosys chairman Nandan Nilekani has said, the approach is essentially: let the major US labs build the frontier models, and India will use them to create synthetic data, build smaller language models, and deploy AI at scale using locally relevant data. India’s AI startup funding remains a small fraction of US investment. Domestic model efforts like BharatGen are focused on Indian-language capabilities for domestic deployment, not frontier competition.

India’s governance approach mirrors this philosophy. The AI Governance Guidelines are principle-based and explicitly pro-innovation. There’s no standalone AI law. The government’s posture is clear: don’t regulate what you want to grow.

The blind spot: the largest sovereignty deficit in the global AI race

India is building the world’s largest AI deployment infrastructure while remaining entirely dependent on foreign models and foreign chips. Indian languages comprise less than 1% of global web content, making frontier model training structurally difficult. Indian venture capital demands commercial returns too fast for the long, capital-intensive model-building cycle. And India’s digital public infrastructure — 800 million health IDs, the Aadhaar biometric system, the UPI payments network — creates an application surface that is unmatched globally but also creates enormous data governance risks if the intelligence layer is controlled abroad.

Indian citizens, for their part, are the most engaged of the three models — but in a paradoxical way. India has among the largest ChatGPT user bases in the world. They are enthusiastic adopters of foreign AI. Civil society, meanwhile, is raising alarms about how AI tools may be used for surveillance and political targeting. India’s public is simultaneously the world’s most eager consumer of AI and among the most exposed to its governance failures.

The Strategic Question

Each model optimizes for a different layer of the stack. The US controls the intelligence layer and the supply chain but faces a democratic revolt over the physical layer. The Gulf controls the infrastructure layer but depends on the US for everything that runs on it. India is building the application layer at a scale no one else can match but has no indigenous control over the models or the hardware beneath it.

The question for infrastructure investors, energy executives, and policy strategists is not which model is right. It’s which model survives a supply chain shock — a chip export ban, an energy price spike, a regulatory reversal. The US model is most vulnerable to its own citizens. The Gulf model is most vulnerable to the USA’s foreign policy. The Indian model is most vulnerable to the models it doesn’t own.

The sovereign AI race isn’t a single competition. It’s three parallel experiments in what “control” means in an era when intelligence is becoming infrastructure. The results will reshape energy markets, capital flows, and geopolitical alliances for the next generation. And none of the three architects has solved the full puzzle yet.

EDITORIAL NOTE

This analysis represents the author’s interpretive framework. The “three models” taxonomy and “blind spot” assessments are analytical constructs, not formal categories used by the governments discussed. Gulf and Indian investment figures reflect headline announcements and government projections; finalized commitments may differ. China’s parallel sovereign AI stack — including domestic chips (Huawei Ascend), foundation models (DeepSeek, Qwen), and cloud platforms — is excluded here because it operates outside the US-anchored supply chain ecosystem analyzed in this piece and warrants separate treatment.

Dell’Oro Group’s latest forecast, published last week, projects the top four US hyperscalers — Amazon, Microsoft, Google, and Meta — will spend close to $600 billion combined this year, up sharply from prior years. Capital intensity across these companies has reached 45–57% of revenue, levels historically associated with utilities, not tech firms. GPUs and custom AI accelerators now account for roughly a third of total data center capex. By any measure, this is the largest infrastructure buildout since the post-war industrial era. But the geography of that buildout has fundamentally shifted over the past twelve months, and most infrastructure strategies haven’t caught up.

The shift traces back to an economics concept from 1865. When DeepSeek released its V3 and R1 models in late 2024 and early 2025, demonstrating that frontier-capable models could be built for a reported $5.6 million in GPU compute costs — a fraction of what Western labs assumed necessary, even after accounting for excluded R&D and infrastructure overhead — the immediate market reaction was panic: if AI is getting cheaper, who needs all this infrastructure?

The answer turned out to be everyone.

The Peterson Institute confirmed it this month: the cost to achieve a comparable score on a challenging AI benchmark plunged from $4,500 per task to $11.64 over the course of 2025. Yet AI usage dwarfed those efficiency gains. Anthropic’s annualized revenue run rate climbed from roughly $1 billion to over $9 billion in twelve months. OpenAI projected $20 billion in 2025 revenue, up from around $3.7 billion in 2024. Google Cloud revenue grew 48% year-on-year. Economists call it the Jevons Paradox — when efficiency drops cost, consumption explodes.

What changed isn’t the scale of the buildout. It’s who’s building and where.

Before 2025, the infrastructure narrative was centralized: US hyperscalers stacking GPU clusters in Virginia, Texas, and Oregon.

After efficiency breakthroughs collapsed the capital threshold for sovereign AI, the buildout fragmented globally.

India launched BharatGen, a government-backed consortium at IIT-Bombay developing sovereign multilingual models, with Sarvam AI targeting Indian-language benchmarks. South Korea declared its intention to become a top-three AI power. Saudi Arabia committed over $5 billion for a new AWS region alongside its sovereign AI fund. Thailand attracted over $10 billion in data center investment pledges in just six months, with commitments from AWS, Google, and TikTok’s parent ByteDance.

The demand profile shifted too.

Training runs happen once. Inference — running models in production — happens continuously, everywhere. As open-weight models like DeepSeek-R1, Llama, and Falcon proliferate, they need hosting infrastructure in Bangkok, Mumbai, Riyadh, and São Paulo — not just Ashburn, Virginia. Even the Stargate Project adapted: what began as an aspirational $500 billion US-centric initiative has evolved toward distributed “Sovereign Stargate” nodes in markets including Norway and the UAE. The architecture of the buildout is following the architecture of the models — distributed, modular, geographically dispersed.

Dell’Oro projects that while the top four US hyperscalers will represent about half of global data center capex by 2030, emerging AI model builders, neo-cloud providers, and sovereign cloud initiatives are growing at significant rates. Over 50 GW of new data center capacity is expected over the next five years.

The buildout didn’t shrink, but moved. For energy executives and infrastructure investors, the next wave of power purchase agreements, grid interconnection requests, and site selection decisions won’t cluster in Northern Virginia. They’ll show up in Thailand’s Eastern Economic Corridor, India’s emerging compute hubs, the Gulf, and a dozen other markets converting efficiency breakthroughs into sovereign compute capacity. The firms still modeling AI demand as a US-centric phenomenon are working with last year’s map.

Sources: Dell’Oro Group (Feb 2026), Peterson Institute for International Economics (Feb 2026), Epoch AI, SemiAnalysis, Reuters, TechCrunch

The declaration is non-binding. There’s no enforcement mechanism. It formalized relationships that already existed. Which is precisely why mainstream media ignored it, and precisely why infrastructure investors should pay attention.

China has spent two decades vertically integrating its tech stack: domestic chip production (SMIC), rare earth monopolies — 70% of global mining, 90% of global refining, and 94% of permanent magnet manufacturing — captive energy supply, and hyperscale data center capacity, all under centralized coordination. The US and its allies cannot replicate that model individually. So they’re doing it collectively.

Pax Silica distributes strategic capabilities across trusted partners. Japan and South Korea anchor roughly 70% of global memory chip production and significant advanced logic capacity, and South Korea’s national semiconductor strategy explicitly targets becoming the world’s second-largest chip power. The Netherlands holds ASML’s monopoly on extreme ultraviolet lithography systems — the only technology capable of producing cutting-edge chips.

Notably, and unsurprisingly, the EU itself is not a signatory, and Under Secretary Helberg acknowledged “real material policy and philosophical differences” with Brussels.

The UK pulled in $23.6 billion in venture capital in 2025 — equal to France, Germany, the Netherlands, and Spain combined — positioning it as a financial engine for frontier tech. Australia provides rare earth processing outside Chinese control.

The Gulf commits sovereign capital at a scale no one else can match: the UAE alone has deployed $148 billion in AI investments since 2024, and the broader Gulf region has pledged over $2 trillion in AI, semiconductor, and infrastructure commitments.

Singapore and Israel anchor AI research hubs and venture financing.

Each country brings what others lack. None can compete with China’s vertical integration alone. Together, they cover the full stack — from mineral extraction through chip fabrication to data center deployment.

Under Secretary Jacob Helberg framed it bluntly: It’s about industrial policy coordinated across democratic governments to ensure no single authoritarian state controls the physical infrastructure of artificial intelligence.

The timing matters. The Trump administration faces persistent criticism that it’s dismantling alliances and accelerating deglobalization. Yet here’s a sophisticated multilateral framework coordinating hundreds of billions in investment across chips, energy, and critical minerals. Pax Silica rewards capability and alignment, not status or proximity.

The agreement landed with almost zero mainstream coverage because it doesn’t fit prevailing narratives. It’s not a dramatic rupture. It’s not protectionist theater. It’s eleven governments acknowledging that AI infrastructure is too strategic to leave uncoordinated and too complex for any single country to secure alone — even the United States.

Capital flows are reorganizing around trusted technology ecosystems. AI supply chains are fragmenting along geopolitical lines the same way energy access did during the Cold War. The blocs are forming now, during the buildout phase. And the logic of these blocs is simple: shared material interest in controlling the triangle of energy, compute, and power. The strategic question is straightforward: who is building relationships inside these emerging ecosystems, and who is still assuming global markets operate independently of alliance structures?

Pax Silica is the first alliance of the AI era. It won’t be the last.

Sources:

• US Department of State - Pax Silica Initiative

• Pax Silica Summit Outcomes - State Department Briefing, Dec 17, 2025

• Semiconductor Industry Association Statement on Pax Silica

Most AI commentary focuses on chatbots, productivity tools, and model releases. This newsletter takes a different view.

I write about what happens when $600 billion in annual capex meets a power grid that wasn’t built for it. When sovereign wealth funds realize AI leverage starts with electricity, not algorithms. When the geography of compute becomes as strategic as the geography of oil.

We are witnessing one of the largest reallocations of capital in modern history. As AI moves from software to infrastructure, the new geography of power is being shaped by electricity grids, data center footprints, and sovereign energy strategy.

Thesis: AI is reshaping government institutions, the energy industry, infrastructure investment, and state leverage. Whoever controls the physical layer of AI — the electricity, the land, the cooling systems, the capital — holds structural power over everyone else renting compute on top of it.

That’s what this newsletter is about.

What You’ll Get

Every week, one strategic briefing:

• One thesis — A clear argument about how AI reshapes energy, infrastructure, or state power

• One chart — Data that makes the point unmissable

• One implication — What this changes for decision-makers

No news roundups. No tool reviews. No 10-point listicles. Just analysis that helps you see structural shifts before they become consensus.

What I’ll Write About

• AI’s real energy cost, and why compute geography is becoming as strategic as oil geography

• How sovereign wealth funds are converting energy dominance into compute infrastructure

• Why climate policy and AI policy are now inseparable

• Data centers as geopolitical assets — location, power, water, land

• Which governments gain administrative leverage through AI, and which fall behind

• Capital flows at the AI-energy convergence

• Grid constraints, interconnection delays, and what they mean for hyperscaler buildout plans

Who This Is For

Energy executives who need to understand how AI demand reshapes power markets and grid planning.

Infrastructure investors modeling data center demand, grid capacity, and the capital required to bridge the gap.

Policy advisors and government strategists navigating national AI strategies where compute infrastructure equals geopolitical leverage.

Corporate strategists who need to anticipate how AI changes their sector before it’s obvious.

If you work in energy, infrastructure investment, policy, or strategy — and you need to see what’s coming before everyone else does — you’re in the right place.

A Few Ground Rules

I write for people who make decisions, not people who read headlines. That means:

No hype. AI is transformative, but most AI commentary is noise. I focus on what matters for infrastructure, energy, and state capacity.

Forward-looking, not futuristic. I won't speculate about what might happen in 2030. I'll show you what's happening now and what it implies for the next 12-24 months based on capital flows, construction timelines, and policy commitments.

No vendor pitches. This isn’t about tools or platforms. It’s about structural forces.

Data over narrative. Every argument gets a chart. If it can’t be measured or sourced, it doesn’t belong here.

What’s Next

Your next briefing arrives this week. I’m working on a piece about Pax Silica — a coalition that binds 11 countries into a shared AI supply chain covering chips, minerals, energy, data centers, and undersea cables. The name blends Pax Romana with silica, the base compound of silicon. The signatories: United States, Japan, South Korea, Australia, Israel, Singapore, the UK, the Netherlands, Qatar, Greece, and the UAE. India is next.

If you find this valuable, forward it to someone who would benefit. If you want to support the work and get monthly deep dives plus quarterly strategy calls, consider becoming a Paid or Founding Member.

Otherwise, just stay on the list. Every weekly briefing comes to you regardless.

Thanks for being here from the beginning.

— Ivan Ferrari

A note on independence: All opinions shared in this newsletter are my own and do not reflect the views of dmg events, ADIPEC, or any affiliated organizations. This is personal analysis, not institutional positioning.