AI data centers might just save rooftop solar — and your electric bill

Have something to say? Lookout welcomes letters to the editor, within our policies, from readers. Guidelines here.

Drive through Live Oak, Scotts Valley or the Westside and you can’t miss it: rooftop solar panels catching the sun on roof after roof. For years, California all but begged Santa Cruz County families to install them, streamlining permits and dangling rebates. 

Thousands of us said yes.

Then the story flipped. 

The same panels the state once championed got recast as the reason your neighbor’s electric bill keeps climbing – a “cost shift” onto everyone else, solar owners were told, that they should feel guilty about.  

For those of us who did put in solar, we have good news: Artificial intelligence just blew a hole in the guilt theory. Santa Cruz County ratepayers should be paying attention, because we’re the ones footing the bill either way.

The arrival of AI data centers creates a new test of the cost-shift argument. 

These facilities will require enormous amounts of new electricity and billions of dollars in new grid infrastructure. If reducing electricity purchases through rooftop solar is considered an unfair burden on other ratepayers, then adding massive new demand to the grid deserves even closer scrutiny.

That comparison matters because it reveals a bigger opportunity: The AI boom could help solve the very grid challenge that rooftop solar has been accused of creating. Data centers need power quickly and reliably. Distributed solar, community solar and battery storage can help deliver it without waiting decades for new transmission projects.

But let’s back up for a minute. 

Here’s the cost-shift argument in plain English. When you install solar, you buy less electricity from Pacific Gas & Electric. PG&E recovers most of its costs volumetrically, by the kilowatt-hour, so when you buy less, it collects less from you and, because state law lets it recoup its costs regardless, more from everyone else. 

Utilities have spent years pointing at rooftop solar as the culprit behind rising bills.

To be fair, there’s a real issue buried in there: fixed costs like poles, wires and wildfire mitigation don’t shrink just because customers buy less power, and someone has to cover them. That’s a legitimate conversation about rate design. It is not, however, a rooftop-solar problem; it’s a math problem that shows up any time anyone buys less electricity. Swap in LED bulbs, attic insulation, a more efficient refrigerator or just a habit of turning off lights, and the identical “cost shift” occurs. 

The scale of that challenge is what makes the AI boom so consequential.

Hyperscale data centers are asking to plug gigawatts of new demand into a grid that’s already stretched thin. A single 500-megawatt facility can trigger hundreds of millions of dollars in new transmission lines, substations and generation. Multiply that across the dozens of facilities California is fielding requests for, and the infrastructure bill, under today’s rules, lands on ratepayers, including those of us who will never see a data center from our porch in Santa Cruz County.

If a homeowner reducing electricity purchases counts as an unfair cost shift, then adding billions of dollars of new data-center infrastructure to the grid should count for a great deal more. 

Data centers don’t just fail to reduce grid costs, they multiply them. 

Major transmission projects take a decade or more to permit and build, timelines that don’t match what AI companies say they need. Distributed solar, by contrast, can go up in two to three years, requires little new transmission and generates power close to where it’s used.

One grid-optimization study found that a distributed clean-energy system could cost $88 billion less than a centralized build-out. A 2024 analysis put the net benefit of existing rooftop solar to California ratepayers at roughly $1.5 billion in a single year, through avoided line losses, deferred upgrades and added resilience during outages. 

Distributed solar isn’t just clean power. It remains the cheapest, fastest insurance policy the grid has against the AI boom.

And, instead of restaging the same tired fight over who to blame for the bill, here’s the more interesting version of that idea. 

We have spent years trying to deploy rooftop and community solar fast enough to matter for California’s climate goals, without ever finding the money or political will to do it at real scale. 

The AI boom might hand us both. 

Data centers arrive with deep capital and a deadline they can’t ignore, exactly what ratepayer-funded programs have lacked. If AI companies want fast, reliable access to California’s grid, they can help finance rooftop solar, community solar and battery storage in communities across the state, no matter how far those communities sit from the nearest server farm. 

Santa Cruz County will likely never host a data center of its own. Under this model, that wouldn’t stop it from sharing in the benefits. 

That’s the premise behind the Distributed Energy Participation Program (DEPP) Act, a legislative proposal we’re developing: no new taxes, no new utility rate hikes, just a voluntary market that channels data-center capital into community solar. 

Santa Cruz residents, like all Californians, already pay the second-highest residential electricity rates in the country. PG&E’s rates have climbed 41% in three years and 101% in the past decade. We cannot build our way through the AI boom on ratepayers’ backs alone, and we shouldn’t have to.

The data-center boom is arriving whether we want it or not. California can keep rerunning the same argument over who to blame for our electric bills, or it can notice that the AI industry just handed rooftop solar its best argument yet. 

Get this right, and the AI revolution could do something almost nobody predicted: make electricity more affordable while putting more solar panels on more Santa Cruz County properties.

That’s a plot twist worth following.

Angela Lipanovich is a clean energy attorney who runs Estriatus Law and co-founded SolarWAVE Action, a nonpartisan research and advocacy group that turns legal and data analysis into California clean energy policy. She previously served as general counsel to a publicly traded solar company and has helped shape state policy protecting solar ownership rights. Reach her at alipanovich@estriatuslaw.com.

Jenny Folkesson is executive director and co-founder of SolarWAVE Action. A computer scientist who got her doctorate at the University of Copenhagen, she spent two decades in data science and machine learning before turning her open-source instincts toward clean energy and coastal and marine conservation research. Reach her at jenny@solarwaveaction.org.