The Trump administration and a bipartisan group of 13 governors announced an agreement urging the PJM Interconnection — the largest electric grid operator in the U.S. — to “…expand the energy supply, safeguard affordability and support rapid expansion of data-center development.” These are commendable goals, but the unfortunate reality is that the shared grid is ill-equipped to power data centers quickly or fairly.

The U.S. electricity sector is a slow-moving maze of regulations, shaped by decade-long transmission approvals, time-intensive interconnection studies for new generators and large new customers, and overlapping layers of state, regional and federal bureaucracy. Neither the 13 governors nor the president can easily order PJM to do anything, nor is more government intervention a recipe for long-term success.

The regulatory thicket surrounding the electricity industry was tolerable when the pace of change was slow. However, with the rise of AI and renewed growth from manufacturing and electrification, we can no longer endure a sclerotic grid.

Here’s the rub: Reforming the entire industry would be messy and slow. Unwinding electricity’s regulatory complexity without adding risk and political uncertainty is a Herculean task that cannot happen quickly enough to relieve today’s crunch. Further, deep reform gambles the fate of the entire country — a common quip is that electricity is just 5 percent of the economy, but it powers the other 95 percent. We must keep the grid reliable and affordable for everyday Americans.

But there is a way to enable the electricity buildout required by modern industry while protecting consumers and guarding against blackouts. We can allow entrepreneurs and private investors to work outside of the existing sector to establish new, private utilities and new grids free from the regulations that dominate the shared grid.

Sen. Tom Cotton’s (R‑Ark.) DATA Act of 2026 offers a consequential fix. It would exempt certain new electricity providers from the federal regulations that apply to the broader grid. The bill would clear a path for new, off-grid providers to move quickly to serve AI and data center demand.

By focusing on AI and data centers, as well as other large industrial loads, these new utilities and grids would serve large and sophisticated businesses that buy products and services from other parties all the time without regulatory oversight. For these customers, buying electricity should be no different. And with no physical connection to regulated utilities, these new utilities could serve new customers without impacting the cost or reliability of the existing grid.

As it stands, federal utility regulations can easily add years to project timelines. Many of these rules are necessary to maintain reliability, since electricity must be balanced in real time, 24 hours a day, seven days a week. Policymakers are reluctant to tinker with the entire grid because the downside is too great — blackouts are hazardous to public health and politically toxic.

Even well-meaning regulations create uncertainty for investors: An investor may think they have a great project, but it can be shot down by regulators or delayed to the point of economic extinction. Worse, the rules of the road can change every four years.

The DATA Act would promote speed and certainty. It doesn’t use taxpayer money or risk burdening existing ratepayers, who have little choice but to buy their electricity from the local monopoly. If entrepreneurs and private investors are willing to risk their money and create an off-grid utility, the DATA Act would effectively say, “good luck and Godspeed!” They would sink or swim based on their ability to meet customers’ needs. The cost of failure — or a potential AI bubble — would hit investors rather than ordinary ratepayers.

The biggest benefit of the DATA Act may prove to be what the electricity sector has lacked for about a century: innovation. For too long, innovation in the sector has moved at a snail’s pace. As our grids became “too big to fail,” reliability — not speed or efficiency — became the objective of the country’s regulatory framework. Unfortunately, innovation can put reliability at risk. 

Entrepreneurs and private investors, on the other hand, embrace risk. Building a new off-grid utility is inherently risky. It will need to be innovative in the hope of outcompeting existing utilities for new business. And without federal regulators poring over the new utility’s proposed innovations, it would be free to try new things. The good and bad ideas alike will be instructive to entrepreneurs and regulators, and the best ideas can be safely imported into the regulated utility space.

Letting AI data centers and other large customers use new, off-grid power providers is not a bet against regulation, regulators, or the regulated grid. It is simply an acknowledgement that a regulatory system built for a 20th century monopoly grid cannot power a 21st century AI economy. Giving entrepreneurs the room to build around that system and create something new and dynamic is the fastest and safest way forward.