Good Rules Make Good Neighbors: Why Oklahoma's Data Center Bill Is a Conversation Worth Having

Let's start with the part nobody in our industry is supposed to say out loud: House Bill 2992 has the right instinct.

The bill, as filed, draws a line around a new kind of electric customer — the data center, the AI campus, the facility that shows up asking for seventy-five megawatts or more — and says that customer should pay its own freight. Separate tariffs. Costs assigned by who causes them. Credit posted up front so that if a facility powers down or pulls out, the bill doesn't drift onto the family running their air conditioner in Ada or Altus. A ten-year commitment, so a company asking a utility to build for it has to stick around long enough to pay for what it asked for.

We could fight that. We're not going to, and here's the honest reason: across the country, the places where our industry tried to move fast and explain later are the places where the projects are now dead. Community opposition stalled or killed roughly $98 billion in data-center development in a single three-month stretch last year. Nationally, 78 percent of Americans tell pollsters they're worried these facilities will push their power bills up. You do not build a twenty-year industry on top of that kind of resentment. You build it on top of rules that let a skeptical public watch the meter and see that the math is fair.

So we support the spine of this bill. The principle the regulators call cost causation — you pay for the infrastructure built to serve you — is one our member companies have already started endorsing voluntarily, pledging to bring or buy our own power rather than lean on the existing grid. HB 2992 simply writes that handshake into Oklahoma law, and a handshake written down is worth more than one that isn't.

That doesn't mean the filed version is finished. A few things deserve a careful second read before this becomes statute. "Cost causation" is a fair principle, but it can be abused as a license to load every contingency onto the new customer, including costs a utility would have incurred anyway — and a tariff that punishes load growth rather than pricing it will simply send investment, jobs, and tax base to the state next door. The ten-year term makes sense for a hyperscale campus; applied bluntly, it could scare off the smaller and earlier-stage projects that often become the big ones. And the credit and stranded-cost provisions need to be calibrated to real risk, not to worst-case theater.

These are details, not objections. We raise them the way a good neighbor raises the question of where the fence goes — before it's built, in daylight, with both deeds on the table. Oklahoma has a genuine opportunity here: to be the state that got the data-center bargain right, where the industry pays its way, the ratepayer is held harmless, and the lights and the jobs both stay on. HB 2992 is the start of that conversation. We're ready to have it.