Thoughts from the Pathward Executive Summit
Written by Shawn Andrews, CEO of BridgePeak Energy Capital
BridgePeak was recently invited to present at the Pathward, N.A. Executive Summit at The Lodge at Torrey Pines in La Jolla, California. The Summit brings together senior executives from across the payments and fintech ecosystem to examine the forces shaping financial services and to explore practical implications for growth, risk management, and partnership development.
One theme kept repeating itself: the role of artificial intelligence. Nearly every presentation made some reference to AI, whether in the context of payments innovation, agentic commerce, regulatory considerations, or operating model transformation.
Following sessions on emerging fintech trends and agentic commerce, my presentation focused on a foundational question: how energy infrastructure and grid capacity impact the ability to scale AI‑driven business models in the U.S.
My session, “Powering AI – The Impact of Energy on U.S. AI Dominance,” examined the infrastructure required to sustain AI deployment at scale, and why energy is becoming the binding constraint.
BridgePeak’s Perspective on Energy and AI
BridgePeak’s perspective on AI‑driven energy demand is shaped by our experience providing capital across the full lifecycle of energy infrastructure. This experience provides a practical lens for evaluating what AI deployment requires in the real world. AI does not operate in abstraction. It operates on physical infrastructure that must be financed, built, and maintained over decades.
AI isn’t just a compute problem
AI is often framed as a compute challenge, but its scalability is equally dependent on energy availability, grid reliability, and infrastructure finance. Without a deliberate strategy to expand and modernize energy capacity, AI leadership becomes difficult to sustain at scale.
AI Is Driving a Step‑Change in Electricity Demand
U.S. electricity demand is projected to rise approximately 40% by 2035, requiring the addition of more than 800 GW of new installed capacity to meet peak demand.
This represents a fundamental break from historical trends. For much of the past two decades, electricity demand grew at roughly 0.3% annually. Current projections reflect growth closer to 3.4% annually between 2023 and 2035—roughly ten times faster than the long‑term historical rate.
Three forces are driving this acceleration:
– Data center expansion driven by AI workloads
– Electrification across transportation and industry
– Manufacturing reshoring and advanced industrial capacity
Among these, AI‑driven data centers are emerging as one of the most capital‑intensive and energy‑intensive components of the modern economy.
Data Centers and the Largest CapEx Cycle of the Millennium
AI is projected to drive the largest capital expenditure cycle this millennium. In 2026 alone, AI‑related capital investment is estimated at $625 billion, surpassing both the dot‑com boom and the shale energy buildout of the 2010s.
This scale of investment has direct implications for the power sector. Data centers require:
– High‑density, always‑on load profiles
– Exceptional reliability and resiliency
– Predictable long‑term power costs
These requirements increasingly influence site selection, financing structures, and regional economic development. Where power is scarce or constrained, AI deployment slows. Where power is abundant and well‑structured, deployment accelerates.
The grid constraint is structural
Despite strong capital markets and technological leadership, the U.S. grid today was not built for this level of demand growth. It’s fragmented, aging, and increasingly difficult to expand at the pace required. Meeting AI‑driven load growth requires optimizing existing infrastructure and building significant new capacity across generation, storage, and distribution.
This is not simply a question of adding megawatts. It is a question of timing, interconnection, dispatchability, and coordination across market structures that vary significantly between regulated and deregulated states.
Distributed energy as a practical solution
One of the most practical near-term solutions is distributed energy.
Technologies such as community solar, commercial and industrial solar‑plus‑storage, utility‑scale storage, and behind‑the‑meter battery systems are increasingly essential to meeting demand growth without over‑reliance on long‑dated transmission expansion.
Battery storage, in particular, has evolved rapidly. Once limited to paired renewable projects, storage now benefits from standalone investment tax credit eligibility, new contract structures, and expanding revenue mechanisms, particularly in deregulated markets such as ERCOT.
From a capital perspective, storage assets are differentiated not by technology alone, but by where they sit on the system and what problem they are designed to solve.
Why this matters now
China’s rapid electrification demonstrates what coordinated investment can achieve. The U.S. has structural advantages, including energy abundance, innovation leadership, and deep capital markets. However, those advantages only matter if they’re effectively mobilized.
My final thoughts are straightforward: U.S. AI dominance is inseparable from energy strategy. Compute capability without reliable power is theoretical. Delivering AI at scale requires confronting energy constraints head‑on and financing the infrastructure that resolves them.
I appreciate Pathward for bringing together leaders who are open to these critical discussions. The intersection of energy and AI is becoming a defining factor in how the next phase of US AI growth unfolds.
Q&A
Q: Why is energy becoming a constraint on AI growth? AI is often framed as a compute challenge, but at scale it becomes an energy challenge. Data centers require large, continuous power loads with high reliability. As AI deployment accelerates, the availability of power and grid capacity increasingly determines how quickly these systems can be built and scaled.
Q: How much is electricity demand expected to grow? U.S. electricity demand is projected to increase by approximately 40% by 2035. This marks a significant shift from the past two decades of relatively flat demand and is being driven by AI data centers, electrification across industries, and the reshoring of manufacturing.
Q: What is the biggest constraint in the current power system? The U.S. grid is fragmented, aging, and slow to expand. Interconnection delays, transmission limitations, and regulatory complexity create structural bottlenecks that make it difficult to meet rising demand.
Q: What role does distributed energy play in solving this? Distributed energy resources such as solar paired with storage and standalone battery systems provide a practical way to meet localized demand growth. They reduce reliance on large-scale transmission expansion and are becoming an essential part of how the system adapts to new load requirements.