Virginia Gov. Glenn Youngkin

Artificial Intelligence policy is quickly spreading into the critical domain of energy.

It’s a two-sided issue. On one hand, AI computations and data storage require a lot of electricity, and its boom is expected to increase global power demand. On the other hand, the technology holds a lot of promise for boosting energy innovation and efficiency.

In this segment, we look at the intersection of AI and energy innovation and what’s on the horizon for the sector as technology moves the world into the future.

Be sure to tune into our accompanying podcast, where we’ll further dissect the issue.

The State of Play

While Washington and state capitals have been focused on questions of safety or accuracy, policymakers are now preoccupied with how the U.S. can power all the computations that AI demands.

Energy needs for U.S. data centers are expected to triple in the next five years. The projected growth is so rapid that AI has become one of the chief motivators behind an array of changes to energy policy; the two have essentially fused. The expansion is a challenge for Washington, which oversees transmission lines and other cross-state issues. But states are often the primary energy regulators, subjecting them to various opinions, options and obstacles.

As the AI industry evolves, regulators and policymakers will continue to examine what types of energy the technology will require, where data centers will go and how the complex dance between lawmakers, state officials, the White House and the private sector takes shape.

Boosting innovation: AI has a lot of potential to change how we consume and produce energy. Algorithmic systems can theoretically predict peak demand times or detect energy-draining events like cyberattacks. AI companies are also eager to use their offerings to push forward scientific discovery, which could lead to advances in batteries, biofuels and other renewable technologies.

AI is also used to manage energy usage, including in massive industrial sites or older commercial buildings. 

Energy for AI: The biggest question now is how exactly policymakers want to expand energy production. Energy being at the center of the AI policy conversation is a new issue — in part due to the results of the election — and Washington is still figuring out its next moves.

For now, few Republicans in power are worried about the climate change effect of expanding power generation, although the efforts could get more complicated if Democrats return to power. 

One of the big questions, though, is whether GOP calls for an all-of-the-above energy strategy will rely heavily on renewables or if energy efforts will focus more on traditional fossil fuels and nuclear power.

Meanwhile, states are busy trying to sign up new projects — either energy plants or data centers — sometimes sponsored by a particular company. Meta has planned investments in Louisiana, and Microsoft plans to restart Three Mile Island in Pennsylvania.

The Spotlight Interview: Virginia Gov. Glenn Youngkin (R-Va.)

To get a perspective on the connection between energy and AI, we spoke with Virginia Gov. Glenn Youngkin.

Virginia is the world’s leading data center market, as Ashburn is widely considered the “data center capital of the world.” AI industry experts often claim that 70% of internet traffic passes through undersea cables running into the commonwealth and through its data centers.

Much of the first internet infrastructure originated in government labs near Washington, D.C., and was then fostered by contractors and commercial vendors setting up to work with the public sector. The defense and intelligence sectors especially need U.S.-based workloads and processing.

Youngkin described some of his priorities for using AI, which he plans to apply to basic services like chatbots that answer FAQs or to streamline DMV functions for residents, while also attempting to get rid of inefficiencies. He also said it’s imperative to have a human in the loop of any decisions.

In 2023, Youngkin issued a directive to the state energy department to “better understand the impact of AI on [its] power generation requirements” due to increasing AI usage. More on that below.

But he also gave insight into his approach to energy, infrastructure and other topics key to building AI. Here are the key takeaways from our chat, which have been edited for brevity and clarity.

The Outlook

Even though the energy-for-AI picture is just starting to come into focus, there is already a wealth of initiatives aimed at moving the government forward. 

Youngkin, for instance, discussed the state energy utilities’ plans to double power generation by 2039, which follows his AI directive. Youngkin also set up an AI task force last fall to weigh in on policy, standards and education. And the legislature is also working on a high-risk AI bill.

On Capitol Hill, the Senate Commerce Committee and the House Energy and Commerce Committee will take the lead on energy and permitting, whether in reconciliation or via a separate vehicle. Both efforts are broader than AI, but the chairs routinely cite the technology as perhaps the chief reason for the urgency of their work.House Energy and Commerce Committee Chair Rep. Brett Guthrie (R-Ky.) has said he wants to start with the House GOP leadership’s energy package. There are bipartisan talks on getting permitting reform going again.

Freshman Rep. Julie Fedorchak (R-N.D.), a former public utility commissioner, has also started a working group on the topic.There are other efforts as well. Sen. Lisa Murkowski (R-Alaska) last year introduced a bill to authorize studies on AI and energy through the Energy Department. That measure could be revived this Congress as Republicans seek to make some big moves in the energy realm.

White House action: In the meantime, the White House announced in mid-February it would create a National Energy Dominance Council. That initiative reflects in part the conception of energy and AI as strategic advantages in international competition, particularly with China.

The council’s mandate is to come up with ways to improve “the processes for permitting, production, generation, distribution, regulation, and transportation across all forms of American energy” and also to boost private investment and cut “red tape.”

The private sector has already started to invest in mega-infrastructure projects, like the collaboration known as Stargate between OpenAI, SoftBank and Oracle.

— Ben Brody