AI Energy Infrastructure, 123 GW U.S. Demand by 2035, $1 T Annual Grid Investment, and 24 Pax Silica Nations (2026 to 2030)

AI Energy Demand Projects: Pax Silica’s 1, 050 TWh Challenge and Grid Modernization Imperative

The exponential growth of artificial intelligence, the central focus of the December 2025 Pax Silica initiative, has triggered an energy demand crisis that is now the primary constraint on data center deployment. The strategic conversation, which from 2022 to 2025 centered on securing semiconductor supply chains, has abruptly shifted in 2026 to securing gigawatts of power. The core mission of Pax Silica, building a resilient AI ecosystem, is no longer limited by silicon fabrication capacity but by the physical limits of global electricity generation and transmission infrastructure.

From Chip Race to Power Race

The AI industry’s focus has pivoted from chip access to power access as the true bottleneck. Before 2026, geopolitical strategy was defined by semiconductor leadership and initiatives like the CHIPS Act. Today, the “energy-for-compute” model is the central pillar of strategy, as the physical infrastructure for power cannot be deployed at the same speed as software innovation. This has created a direct confrontation between digital ambition and physical reality, a core tension driving the U.S. energy infrastructure boom in 2026.

The Scale of the Energy Deficit

The electricity demand from AI is outstripping existing grid capacity and growth forecasts. The evidence points to a systemic challenge that requires immediate, massive investment to avoid stalling the AI buildout.

• Global data center electricity consumption is projected to reach 1, 050 TWh by 2026, an amount comparable to the entire energy consumption of Japan.
• In the U.S. alone, data centers consumed 4.4% of national electricity in 2023. This figure is on track to triple by 2028, with Deloitte projecting AI-specific data center demand could surge from 4 GW to 123 GW by 2035.
• To meet 2030 climate and energy targets, the world must build more renewable capacity between 2026 and 2030 than in all prior years combined, necessitating nearly $1 trillion per year in grid investment.

Corporate Strategy Shifts to Power Acquisition

Major technology companies are now behaving like energy firms, making multi-billion-dollar investments predicated on securing power. Google’s $15.1 billion commitment to AI infrastructure in Finland was driven by the region’s abundant wind power. Similarly, Open AI’s $500 billion “Stargate Initiative” and other hyperscaler plans are forcing a power-first reassessment of AI infrastructure, where memory and compute are now secondary to energy availability.

Pax Silica Investment: $5 Trillion Directed at AI Power and Grid Infrastructure

The multi-trillion-dollar investment forecasts for AI are increasingly being channeled away from pure compute infrastructure and toward the underlying energy generation and grid systems required to power them. Financial institutions have recognized that the AI revolution cannot be financed with software-level valuations alone; it must be underwritten by hard asset investment in power plants, transmission lines, and critical mineral supply chains. This shift is evident in new capital allocation strategies and dedicated industrial funds launched in late 2025 and 2026.

JPMorgan’s $1.5 T Security Initiative

Financial institutions are mobilizing capital specifically for these strategic industries. In October 2025, JPMorgan Chase launched a $1.5 trillion “Security and Resiliency Initiative” explicitly targeting investments in sectors critical to national security, including minerals, energy, and logistics. This initiative includes a $10 billion direct investment allocation to strengthen critical mineral supply chains, a direct enabler of both semiconductor and energy hardware manufacturing.

The Trillion-Dollar AI Buildout

Forecasts for total capital expenditure are astronomical, reflecting the scale of the infrastructure required. Financial models from major institutions quantify a historic capital cycle focused on the physical layer of AI.

• JPMorgan estimates that over $5 trillion will be spent on global data center and AI infrastructure, including power, in the five years following November 2025.
• Pw C projects global investment in AI infrastructure will reach a cumulative $31.6 trillion through 2050, underscoring the long-term nature of this buildout.
• In the U.S. alone, total investment in data centers and AI infrastructure is projected to reach $10.3 trillion between 2025 and 2032.

Table: AI Infrastructure and Energy Investment Commitments (2025-2026)

Pax Silica Alliances: 24 Nations Target Minerals and Energy Security

The Pax Silica initiative and its critical minerals counterpart, FORGE, represent a strategic realignment where international partnerships are explicitly formed to secure the physical inputs for AI. This has expanded beyond the original semiconductor focus to encompass the entire supply chain, from rare earth elements essential for chips and magnets to the copper and lithium required for grid-scale energy storage and transmission.

FORGE Mineral Alliance

The Forum on Resource Geostrategic Engagement (FORGE), launched in February 2026 as the successor to the Minerals Security Partnership, formalizes this effort. It aims to create a preferential trade-and-investment zone to “friend-shore” mineral supply chains away from high-risk sources. The coalition includes the United States, Australia, Canada, Finland, France, Germany, India, Japan, South Korea, and the United Kingdom, among others.

Bilateral and Corporate Partnerships

This multinational framework is reinforced by direct bilateral and corporate actions. The U.S. signed a Strategic Critical Minerals Cooperation Framework with India in May 2026 to deepen cooperation in mining and processing. On the corporate front, JPMorgan is helping finance Korea Zinc’s $7.4 billion critical minerals processing project in Tennessee, a direct outcome of this strategic alignment. Power infrastructure companies like LS Electric are securing major DC power contracts to support the data center buildout within this framework.

Table: Key Pax Silica and Related Alliances

Grid Infrastructure Maturity: Legacy Power Systems Challenge the 2026 AI Buildout

While AI models and semiconductor technology like that from Applied Materials are advancing exponentially, the grid infrastructure required to support them remains a critical bottleneck. The technology exists, but its deployment maturity, scale, and speed are misaligned with AI’s demand curve. The focus of technological risk has shifted from the data center to the substation, with the slow-paced modernization of the grid becoming the single greatest inhibitor to the Pax Silica vision.

The 2022-2025 Technology Focus

In the period leading up to 2026, technology maturity discussions were dominated by advancements in the digital realm. The primary concerns were centered on semiconductor fabrication, with progress measured in nanometers (e.g., the race to 3 nm nodes), advanced chip packaging, and the architectural complexity of large language models. The underlying power infrastructure was largely assumed to be a given.

The 2026 Grid Reality Check

Since the start of 2026, the conversation has been forced to confront physical reality. The maturity bottleneck is now in grid-level hardware and systems integration. This includes decade-old electrical transformer designs with manufacturing backlogs stretching over two years, cumbersome permitting processes for new high-voltage transmission lines, and the challenge of integrating intermittent renewables at massive scale. This reality is forcing utilities like Enel to accelerate grid modernization with digital twins and robotics.

Alternative Power Sources at the Forefront

The inadequacy of traditional grid expansion is bringing alternative and distributed power sources into the commercial spotlight. Technologies that were once considered niche are now being evaluated for at-scale deployment to power data centers directly. This includes a new look at the commercialization of fuel cells, as well as small modular reactors (SMRs), enhanced geothermal, and long-duration energy storage. The technological maturity of these solutions is now a matter of national strategic importance, not just commercial viability.

SWOT Analysis: Pax Silica Energy and Infrastructure Risks

The Pax Silica initiative’s strength lies in its unified geopolitical alignment and massive capital mobilization for the AI revolution. However, this strategic push confronts a critical weakness in the physical constraints of global energy infrastructure. This creates a complex dynamic of opportunities for new energy technologies and significant threats from potential execution failures, supply chain bottlenecks, and the immense challenge of grid modernization highlighted by programs like the U.S. GRIP for grid infrastructure.

Table: SWOT Analysis for Pax Silica’s Energy-for-Compute Strategy

2027 Scenario: Will Pax Silica’s AI Ambitions Be Curtailed by Grid Capacity?

The most critical indicator for the success of the multi-trillion-dollar AI build-out in 2027 will not be semiconductor sales, but the rate of new grid-scale power generation and transmission projects achieving final investment decision. This is the longest-lead-time item and the ultimate physical constraint on the Pax Silica vision. The strategies of AI firms like Mini Max are now built around mitigating this energy cost.

• If this happens: A continued slowdown in utility-scale renewable PPA signings, further delays in SMR licensing, or persistent two-year-plus backlogs for high-voltage transformers…
• Watch this: The stock performance of regulated utilities versus hyperscalers, announcements of data center projects being “paused” or relocated due to power scarcity, and the cost of capital for new energy projects.
• These could be happening: A market bifurcation where companies with direct access to power accelerate, while others stall. Geopolitical tensions could shift from disputes over chips to competition for energy resources, skilled electrical engineers, and grid hardware manufacturing capacity.

Global Tech Standards Bifurcated Between Pax Silica and WAICO Blocs
The global map clearly delineates two distinct geopolitical blocs: Pax Silica (blue) and WAICO (red), with almost entirely non-overlapping memberships. North America, Europe (including the EU and its individual members), and Australia align with Pax Silica, while China leads the WAICO bloc, with minimal countries like Kazakhstan being signatories to both.

(Source: The Seconded European Standardization Expert for China project; U.S. Department of State — via The State of AI Global Governance and Its Implications for the U.S.-China Summit)

The questions your competitors are already asking

This report covers one angle of the energy constraint on AI development. The questions that matter most depend on your work.

• Data centers building their own power supply
• New grid transmission projects in the US
• US electrical transformer manufacturing capacity
• Critical mineral projects getting funding in the US

This report does not answer these. Enki Brief Pro does.

Your question, your angle, your framework. SWOT, PESTL, scenario modelling. The same niche depth, built around the decision your work actually depends on.

Run your first brief in Enki Brief Pro