AI’s Energy Demand and Economic Implications

Artificial intelligence is reshaping much more than just jobs and international relationships; it’s also becoming one of the most energy-intensive technologies we’ve seen. It’s pretty alarming, really. With the power grid already under strain, AI is making things even more complicated.

According to recent estimates from the International Energy Agency, global data centers are projected to more than double their electricity usage by 2026, surpassing 1,000 terawatt hours. This surge isn’t just due to the sheer energy needed to run these operations but also reflects the immense computational requirements for training and deploying AI models.

Utilities across the U.S. are flagging concerns about a lack of capacity. To make matters worse, various regulatory hurdles, along with climate objectives, have made it quite challenging to bring new energy sources online. Essentially, we’re falling behind in meeting the energy demands of an AI-driven world.

So, what happens when the national agenda clashes with local economic realities? We often see backlash over state-level environmental, social, and governance (ESG) issues that frequently get overlooked.

While the intent of ESG investing was to promote corporate accountability, it has also imposed new limits on financing for energy and infrastructure development. When major financial institutions cut back on financing for oil and gas companies, some states responded by passing laws that prohibit public contracts with those firms.

Texas was among the first to take action in 2021, pushing through legislation that favored local bond underwriters like JPMorgan and Citigroup. Critics warned this could escalate borrowing costs for local governments. However, a recent study showed that, contrary to predictions, the exit of ESG-sensitive companies did not significantly affect pricing.

This Texas policy did not lead to a systematic rise in borrowing costs. Interestingly, Oklahoma’s similar policies enacted in 2023 yielded comparable results.

What accounts for this unexpected outcome that some ESG advocates anticipated? It partly reflects a long-term shift in the financial structure of cities. Over the last two decades, underwriting spreads have decreased, and competition among states—especially those with zero income tax—has created robust demand for investors.

Moreover, markets often adapt when states challenge perceived overreach by businesses and rating agencies. Other underwriters step in, investors adjust, and life goes on.

This is a significant lesson, especially given how AI’s energy demands require similar reassessments. National climate objectives sometimes prioritize decarbonization while traditional energy sources struggle to keep pace before alternatives are viable. Meanwhile, investment in green technology is surging—though perhaps at the cost of system reliability.

Training one advanced AI model can use several gigawatt-hours of energy. The energy needed for running these models, especially at scale, can outstrip that of training in the long run. To tackle this, we need to focus on energy efficiency and grid flexibility, ensuring that power flows can adapt to shifting demands across different regions and time zones.

The National Energy Control Council, set up by President Trump, is a practical approach to bridging the gap between ambitious national goals and local implementation. By coordinating state and private interests, the council aims to speed up permit timelines and streamline the regulatory processes that often hinder energy infrastructure development.

This council also encourages more transparent negotiations between utilities and large corporate energy consumers, fostering investment in renewable energy while ensuring the grid’s reliability. Instead of enforcing top-down mandates, it promotes incentive coordination and addresses long-standing bottlenecks.

In this evolving landscape, states need to be proactive. They must safeguard their energy futures to meet the demand from AI and stay economically competitive. Some states are already taking steps in this direction. For example, Arkansas is fast-tracking natural gas permits, Georgia is investing heavily in nuclear energy and modernizing its grid, and Texas continues to lead in wind energy while boosting natural gas production for stability, especially after the disruptions from the 2021 winter storm.

Some may worry that these actions conflict with wider climate goals. However, many agree that alternatives—such as unstable power grids and rolling blackouts—are far worse. Policymakers need to understand that reliability and capabilities are not simply optional in today’s digital economy. Moreover, pursuing energy neutrality is not a sustainable long-term strategy.

If ESG financing has taught us anything, it’s that when national trends threaten local interests, states can push back. The market responses to these interventions aren’t always disastrous; sometimes they’re neutral, or even beneficial, by tweaking an approach that misaligns incentives.

As we transition to new energy sources and AI accelerates that change, it’s crucial to rethink how power is balanced between state interests and federal or corporate plans. We simply can’t afford the luxury of ignoring potential AI-related power outages.