Can Oil And Gas Solve The AI Power Dilemma

Authored by Joe Brettell via UtilityDive.com,

The promise, peril and possibilities of artificial intelligence continue to capture the cultural and business zeitgeist worldwide. Hardly a conference or long-form interview can be held these days without a panelist or pundit commenting on the technology’s implications for their profession.

Yet despite being the hottest topic in every circle, AI’s ultimate challenge isn’t technological but physical. After years of breathless speculation and prediction, the issue remains the same: AI needs more energy.

Data center power consumption

Amidst this backdrop, the oil and gas industry faces a similarly fundamental challenge: a shifting production frontier and evolving path to continued growth. After a decade of efficiency-driven growth, the era of easy barrels is waning. Diamondback Energy CEO Travis Stice captured the new reality in a recent letter, warning of the increasingly dim prospects for expanding production amid geological constraints and rising costs. Other energy majors have issued similar cautions, a sharp departure from the boom years of the shale revolution when abundant, low-cost reserves, followed by shareholder-focused production, made the industry a market favorite.

Now, with resource intensity rising, global volatility accelerating and economic conditions tightening, the industry is under pressure to find its next value horizon.

That horizon may be converging with AI.

The pairing makes increasing sense. While initially circling one another warily, major players in energy and technology have become increasingly intertwined. At major gatherings like CERAWeek, energy executives and tech leaders now share the same stage — and increasingly, the same strategic questions. How do we scale the infrastructure to match exponential AI growth? Who will supply the energy to power it? And how do we do so fast enough while dealing with rising environmental, social and regulatory concerns?

These challenges come amid a stark reality: AI’s computational appetite isn’t just increasing — it’s exploding. Several recent studies demonstrate that power demand will soar by the end of the decade, presenting real challenges for utilities and their customers who are already grappling with rising costs.  

That creates both a dilemma and an opportunity. As federal and state incentives for clean energy projects face legal and political headwinds — even amid substantial private investment — the timeline to deliver renewable power at scale is getting longer. Grid interconnection queues, permitting delays and community opposition remain real barriers. At the same time, nuclear and geothermal technologies hold promise, but even under the best-case scenarios, their rollout will take years to materially shift supply.

Which brings us (again) to the topic of natural gas.

Few would dispute that a diverse portfolio of renewables, firm power, storage, nuclear and emerging technologies must meet long-term AI energy needs. But without a tectonic shift, an “all of the above” solution is no longer the political reality.  Natural gas is abundant, dispatchable, and backed by a sector with proven experience in infrastructure delivery, supply chain integration and stakeholder engagement.

Granted, natural gas has its share of controversies. Building new pipelines has become increasingly complex, with communities hostile to natural gas infrastructure and deployment nationwide. Yet, despite these challenges, Exxon and Chevron have already announced serious interest in powering data centers. This partnership is not simply one of convenience but of practicality. It is not about reviving old debates but utilizing practical solutions to solve deeper issues for two pillars of the American economy.

The bottom line is that natural gas offers a workable solution for technology companies racing to deploy AI capabilities and energy companies looking to maintain shareholder value amidst a transitional time in the sector. With nuclear and geothermal both gaining political and investment momentum, gas is unlikely to be a permanent panacea but a critical bridge across a widening gap (yes, the old “bridge fuel” talking point is yet again en vogue).

This convergence between oil, technology, and ultimately, utilities isn’t simply a tactical alignment of convenience; instead, there’s a more profound structural shift — energy and compute are no longer parallel industries but mutually dependent pillars of modern innovation.  If you need evidence, look no further than the recent announcement that Open AI and the United Arab Emirates will open a massive new data center in the country by 2026. Even traditional oil powers are hedging their bet and looking to participate in the changes AI will bring.

However, amidst the race to satisfy shareholders, inventors and policymakers, both industries would do well to remember customers. With concern about AI technology continuing to linger and economic challenges only growing, the political and social environment is ripe for a full-throated pushback from households already frustrated by rising energy bills, service disruptions and increasing skepticism toward unchecked tech expansion in their communities and states.  

Many companies are already making significant strides on this front, with investment in local communities, building dialogue, relationships and trust. Yet just as AI’s technological promise can be limited by something practical like where to plug it in, this growing union between energy and technology sectors can be thwarted by unhappy voters. 

Ultimately, the moment demands coordination and innovation, not competition. Only with pragmatic collaboration between energy developers of all kinds, grid operators and the communities where they operate can we build an energy strategy as dynamic as the technology it supports. In the end, like those paradigm-shifting endeavors, the future of AI won’t be decided by what’s possible in silicon. It will be determined by what’s deliverable in steel, concrete and kilowatt-hours.