AI is dominating energy coverage. That doesn’t mean it’s the whole story.

Aug 18, 2026

Sterling Media Read · Energy, Power and Utilities · July 2026

When the news cycle is crowded, where experts can add to a story, and what makes a source useful

Sterling analyzed 533 July 2026 articles to understand what’s driving energy coverage and where companies may still have something useful to contribute. We looked at the subjects getting attention, how different outlets are approaching them and what kinds of evidence can help a source add to the story rather than repeat it.

AI and data centers turned up in 39% of the energy coverage we sampled. The grid, transmission, and interconnection turned up in 53%. But the more useful number is the overlap: 57% of AI stories were also grid stories.

AI demand is already a crowded angle. What matters is what it changes: how much capacity gets built, who pays for it, and what lands on a customer bill.

 

Energy is moving into the political conversation

Data centers and power costs aren’t staying on the energy beat. In July, they showed up in coverage of a Michigan Senate primary, the political fight over utility bills and data centers in Ohio, and legislation in New Jersey over who should pay for the grid costs created by large new loads.

The questions change when a story travels. Not how much power is needed or what gets built, but who pays, who benefits, and what should be done about it. Fewer people can answer those, and more people are asking.

That is the opening. What a policy actually changes. How costs get allocated. What happens on the grid when a large new load connects. Right now, a specific answer to one of those is worth more than being the biggest name in the room.

 

Where experts can add to the story

THE STORYLINE:
Demand grows → grid constraints emerge → regulators face questions → costs become an issue → politicians notice

Coverage is not spread evenly across that chain. Nearly seven in ten stories touched demand, grid constraints or both. Those are the crowded parts of the conversation.

Further down the chain, coverage thins. Rates, cost allocation, utility operations and the consequences of policy decisions received less attention. Those subjects need someone who can explain what happens next, and fewer companies are offering them.

 

Four cards showing under-covered energy subjects and their share of July coverage. Rates and affordability, 12 percent: thin relative to its political importance; bill impact or cost allocation would add specificity. Water and cooling, 9 percent: a small part of the data-center story; site-level consumption or a siting decision shaped by water would add evidence. Utility operations and modernization, 6 percent: one of the least-covered subjects; deployments and measurable results move past broad claims. Policy and permitting, 25 percent: crowded but expertise-driven; the opportunity is explaining what a proposed rule changes and who it affects.

Four subjects where the coverage is thinner than the stakes. Percentages show each subject’s share of the 533 articles analyzed.

 

Size is not the filter here

No major company owned the conversation. The hyperscalers, large utilities and equipment manufacturers each appeared in less than 2% of the coverage we sampled. The most visible organization was a grid operator, not a vendor or technology company.

Attention is not flowing to the biggest names by default.

WHAT TO DO WITH IT:
Don’t benchmark visibility against the largest names in the sector. Benchmark it against the question you can answer better than they can.

The same story looks different depending on who’s covering it

The same energy story gets covered differently depending on the outlet. T&D World leaned toward reliability, E&E News toward policy and regulation, and Canary Media toward renewables and storage. Business and national outlets brought still different lenses.

Coverage follows the reporter who wants that part of the story. Knowing which part is which is most of the work.

 

Table comparing six outlets by subject emphasis and what each wants from a source. T&D World: reliability and weather in 50 percent of its stories against a 24 percent average; wants to know what held up and what didn't. E&E News: policy and regulation in 52 percent against 25 percent; wants to know what a rule change actually does. Canary Media: solar, wind and storage in 54 percent against 19 percent; wants to know what got built and where. Forbes: AI and data centers in 62 percent against 39 percent; wants to know what it means for a business reader. Bloomberg: grid coverage in 47 to 62 percent depending on the desk; wants to know where the money goes. The New York Times: grid in only 14 percent, the lowest in the set; wants to know what it means for a household bill.

Six outlets covering the same month, each asking a different question. Percentages compare an outlet’s coverage of a subject against the average across all 533 articles.

 

What makes a source useful

The subjects above are crowded with opinion and short on evidence.

There are two ways to provide it.

 

Bring receipts

Move beyond predictions and talking points. Show what actually happened: what a project cost, how long something took, what changed after deployment, what customers experienced, what worked and what didn’t.

Evidence turns a point of view into something a reporter can use.

 

Or explain what others can’t

Plenty of companies can’t release proprietary figures, and they don’t need to. Energy stories increasingly involve systems that few readers understand end to end.

Why do costs move from one customer group to another? What actually slows a project between approval and operation? What happens when a large new load connects to the grid? How does a policy change affect what gets built, who pays, or how long it takes?

Being able to explain how something works, why it matters and what happens next can make an expert useful to a reporter, even without a new number to share.

Bring evidence. Or explain what others can’t. Give the story something it didn’t have before.

Pressure-test your story

  • Where do we have evidence, not just an opinion?
  • What can our experts explain that is difficult to understand from the outside?
  • Which reporters would care, and does our story answer their question or ours?

 

Methodology

Sterling analyzed 533 English-language editorial articles published between July 1 and July 31, 2026, across power, utility and energy trades, business media and national outlets. The analysis was completed and published in August 2026. Stories were coded by subject; subjects overlap, so one story can sit in more than one category. Percentages describe the sampled set and are used to compare subjects against each other, not to size the market. Tone reflects automated scoring and is directional.