AI & Energy: The Connection Brief

The Scale of AI's Energy Appetite

Global data center electricity consumption reached approximately 415 TWh in 2024 — about 1.5% of total world electricity use — growing at 12% per year over the last five years.¹ The IEA projects worldwide data center demand will more than double by 2030 to around 945 TWh — slightly more than the total electricity consumption of Japan — with AI as the most significant driver.³

Gartner's estimate runs higher: worldwide data center electricity demand rising from 448 TWh in 2025 to 980 TWh by 2030, a 16% increase in 2025 alone.² AI-optimized servers are projected to account for 44% of data center power consumption by 2030, up from roughly 21% in 2025.² The IEA frames it more starkly: electricity demand from AI-optimized data centers is projected to more than quadruple by 2030.³

The Grid Under Pressure

Rising data center electricity demand is raising the risk of blackouts across a wide swath of the U.S. during extreme weather events. The supply growth that data centers need has not kept pace with the demand already added.

NERC issued a Level 3 Essential Action Alert — its highest category — after observing that computational loads connecting to the bulk power system produce sudden large load reductions and significant oscillations that occur in seconds, leaving little or no room for real-time response.⁴ The alert outlines seven corrective actions registered entities must implement by August 2026.⁴

PJM Interconnection, the largest U.S. grid operator, procured 145,777 MW in its 2027/2028 capacity auction — 6,517 MW below the reliability requirement.⁵ That shortfall lands on a system serving 67+ million people.⁶

Planning itself has become difficult. Grid operators face deep uncertainty about which announced data center projects will actually materialize, making long-range capacity planning extraordinarily challenging.

The Nuclear Renaissance — Fission Returns for AI

Every major hyperscaler — Microsoft, Amazon, Google, Meta — has signed nuclear power commitments totaling over 9 GW of capacity.⁷ The deals span reactor restarts, existing plant offtake, and first-of-a-kind SMR deployments. The first nuclear-to-AI electrons are expected by 2027.⁷

Fusion Energy Reaches Construction Phase

Fusion energy has moved from theory to construction. Two companies are building real hardware on real timelines, backed by hyperscaler capital.

Helion Energy, backed by Sam Altman, hit 150 million degrees plasma temperature — a milestone that validates the physics.^12a Microsoft has already signed a PPA to purchase electricity from Helion's first fusion plant, scheduled for 2028.^13a Commonwealth Fusion Systems is constructing a 400 MW ARC reactor in Virginia, with power purchase agreements from Google and Eni targeting first power in the early 2030s.^14a

Data Centers in Space

AI's energy appetite is outgrowing the planet's infrastructure. The response: move compute off-world. Six companies — Aetherflux, Axiom Space, Kepler Communications, Planet, Sophia Space, and Starcloud — are building AI infrastructure for orbit on NVIDIA platforms.¹⁵ Google is exploring the same frontier with Project Suncatcher, a moonshot equipping solar-powered satellite constellations with TPUs and free-space optical links to scale machine learning compute in space.¹⁷

The scale ambition is staggering. Terafab targets 1 terawatt per year of output — double the current annual U.S. electricity consumption.¹⁶ In orbit, data centers receive more solar energy than on Earth, with no land constraints and no grid to overload. The math that doesn't work on the ground starts to work in space.

Community Opposition — $64 Billion Blocked or Delayed

$18 billion in data center projects have been blocked and another $46 billion delayed over the last two years, driven by opposition from residents and activist groups.¹⁹

The resistance is not irrational. A single large data center can consume up to 5 million gallons of water per day — equivalent to the annual usage of a town of 10,000 to 50,000 people.²⁰

The water footprint is scaling with demand. The IEA estimates global data center water consumption at roughly 560 billion liters per year, potentially rising to 1.2 trillion liters by 2030 — equivalent to the annual consumption of more than four million U.S. households.²¹ Communities that host data centers absorb the infrastructure strain. The ones that block them absorb the economic loss.

The Efficiency Counter-Narrative

The demand narrative is real. The efficiency counter-narrative is also real. Research is demonstrating that proper model selection alone can meaningfully reduce AI's energy footprint — choosing the right-sized model for a given task rather than defaulting to the largest available.²² Inference optimizations can reduce energy use by up to 73% from unoptimized baselines.²³ Model compression and knowledge distillation deliver ~60% faster inference with ~40% fewer parameters while retaining ~97% of baseline performance. Quantization yields up to ~50% energy reductions. Specialized hardware — TPUs, neuromorphic chips — and data center measures like advanced cooling and virtualization compound the gains further.²⁴

But usage is scaling faster than efficiency can offset. Google processed 9.7 trillion tokens per month two years ago. Last year, 480 trillion. Today: 3.2 quadrillion per month — a 330x increase in two years.²⁵ Uber burned through its entire 2026 AI budget in four months, prompting its COO to question whether the spend is worth it.²⁶ Efficiency buys time. It does not eliminate the demand curve.

The Hyperscaler Energy Arms Race

Four companies — Meta, Amazon, Google, and Microsoft — were responsible for 49% of all global corporate clean energy buying in 2025.²⁷ In the U.S., the concentration is even sharper: these four signed 16,777 MW of corporate renewables contracts, roughly 80% of the total.²⁸ Meta and Amazon led globally, contracting a combined 20.4 GW including 4.7 GW of nuclear power.²⁷

Yet overall corporate clean energy buying fell in 2025 after nearly a decade of growth.²⁷ The market is splitting: hyperscalers are accelerating while the broader corporate universe is pulling back. And the clean energy commitments mask a harder truth — many new digital infrastructure facilities are already relying on fossil fuels to operate, compensating with virtual PPAs and renewable energy credits rather than direct clean power.²⁸

The Insurance Architecture Challenge

Capex is spiking.³⁶ The data center construction market is growing.²⁹ Project values are increasing. Concentration risk becomes more apparent with data center clustering within 20-mile radiuses.³⁰ A single regional catastrophe can hit a high density of insured value simultaneously. When damage occurs, business interruption, contractual penalties, and environmental liability can dwarf the physical loss.^{31, 32}

The market is adapting. Zurich launched Data Center Project Guard — builders risk covering climate control failure, off-site storage, post-construction BI, and weather parametric triggers.³⁴ AIG built a multi-line lifecycle program spanning environmental, professional, D&O, political risk, marine cargo, builder's risk, operational all risks, cyber, and general liability.³⁵

Legal advisors such as Covington mapped the full insurance coverage stack a data center demands across its lifecycle.³³

Insurance brokers and risk advisors also take that lead.

Nine connections. One pattern: AI and energy are in a feedback loop. AI drives unprecedented energy demand. Energy constraints shape where, when, and whether AI infrastructure gets built. Communities are blocking $64 billion in projects. Grids are issuing their highest alerts. Insurance markets are writing policies for facilities whose loss profiles are still emerging.

And the capital commitments keep accelerating — $700 billion+ in 2026 hyperscaler capex, nuclear deals measured in decades, fusion plants breaking ground, AI models trained in orbit.

Whether this produces a better energy system depends on whether the speed of AI buildout is matched by the infrastructure, societal diffusion, and financial markets that have to absorb it.

That is a risk management question — and it is one side of a larger system.

The Full Triad — AI · Energy · Climate

This is one system. The risk question — who carries the risk, and where does it travel? — applies across every connection. The capital decisions happening now in AI infrastructure, energy procurement, grid investment, and insurance architecture will shape the determination whether this system produces a livable planet or an increasingly fragile one.

That is the question Clean Power Whisperer exists to help answer.