Every published Top AI Stories item tagged with Utilities, newest first.
The Environmental Protection Agency has proposed letting state and local agencies decide whether to hold public comment on air permits for large data centers, and separately would let developers start building before permits are approved. Capital B reports that the rural South, where much of this construction is concentrated, has a disproportionately Black population and residents who already learn of projects late because local governments sign nondisclosure agreements with developers. One South Carolina campus under debate spans the area of about 1,200 football fields.
Governor Maura Healey signed an executive order on September 8 barring state agencies from permitting any data center above 25 megawatts of peak demand unless the developer has already secured a community benefits agreement. Developers are expected to bring their own clean generation; those who cannot must pay into a new Ratepayer Protection Fund, with the state environmental agency directed to set that compliance payment by the end of the year. The order also blocks non-disclosure agreements between state agencies and developers, which is the provision aimed squarely at how these deals have typically been negotiated. TechCrunch counts Massachusetts the third state in three months to constrain data center build-out, after Texas ordered audits and New York halted projects above 50 megawatts.
On the same day that it shipped Astra, OpenAI committed $1 billion in subsidized access to its Daybreak cyber models, training and support, targeted to be consumed within six months. It is aimed at water and wastewater systems, electric grid operators, state and local governments, community and regional banks, nonprofits and open-source maintainers, and includes a pilot with the Multi-State Information Sharing and Analysis Center plus more than 35 partner-operated products. This is subsidized demand for OpenAI's own products rather than grant funding, and it lands days after the company said Astra crossed its Critical cybersecurity threshold.
Draft offering documents reviewed by the Wall Street Journal show that SB Energy, the SoftBank-controlled power developer, granted OpenAI warrants valued at $3.6 billion when they were awarded in January — a position worth $5.5 billion by the end of June. What that bought was OpenAI as anchor tenant of the PORTS-Pike campus in Ohio, the site Nvidia agreed to backstop in mid-August. SB Energy is preparing an offering of $5 billion to $7 billion as soon as September, and reports a contracted backlog above $400 billion despite operating no finished data centers yet.
Elon Musk confirmed that SpaceX is adding a blades-and-vanes foundry at its Bastrop, Texas site, where the company bought roughly 830 acres between March and June. Musk says casting those parts is the limiting factor in gas turbine production, and that doing it in-house could pull as much as 18 months out of the wait to bring new turbines online. SpaceX is targeting about 10 gigawatts of AI computing capacity by the end of 2027 — and TechCrunch notes the same turbines carry a measurable local air-quality cost, a fight xAI is already having in Memphis.
The California Energy Commission is putting $8.5 million behind a San Diego Supercomputer Center demonstration of a solid-state transformer that converts medium-voltage alternating current straight to 800-volt direct current in one device, removing the conversion stages that waste power on the way to a rack. Alderbuck Energy supplies the hardware, Emerald AI the orchestration software and San Diego Gas and Electric the grid connection; the two-megawatt build targets roughly 25 percent energy savings. The argument is that a data center can be something the grid leans on rather than a load it absorbs.
Reuters reported that Nvidia has bought a minority position in Cloverleaf Infrastructure, a company founded in 2024 that sits between utilities and developers to secure grid power, cooling and sites. Terms were not disclosed, though the Wall Street Journal put the figure at several hundred million dollars. Cloverleaf will run Nvidia's DSX platform to choose sites and size power. It lands days after Nvidia committed $1.5 billion to SoftBank's SB Energy, and it marks where the chip company now believes the real constraint sits.
TerraPower's Natrium design pairs a 345-megawatt sodium-cooled fast reactor with gigawatt-scale molten salt storage, so the reactor never throttles down: surplus heat goes into the salt and comes back as extra steam when a data center's load spikes. The first plant is under construction in Wyoming under a Department of Energy cost-share worth up to $2 billion, and a second breaks ground in 2027 for an unnamed data center customer. Meta agreed in January to buy eight Natrium plants. The storage was designed for intermittent renewables rather than for AI; the fit with data center load was a happy accident.
Nvidia is investing $1.5 billion in SB Energy, the SoftBank-backed developer of the PORTS-Pike campus in Pike County, Ohio, and is providing credit support covering the land, power and buildings so the site can open at just over four gigawatts and scale to eight. SB Energy will own and operate the data center under a 20-year lease to OpenAI, with Nvidia named the exclusive compute supplier. SB Energy and SoftBank say they will build at least 10 gigawatts of new generation and put at least $4.2 billion into regional grid infrastructure through a partnership with the utility AEP Ohio that is designed to protect ratepayers.
The energy research firm Noreva expects natural gas to pass $10 per million British thermal units at some delivery hubs in the coming years, against a Henry Hub price now just under three dollars. Meta has committed to a 7.5-gigawatt plant in Louisiana and Amazon to a 7.6-gigawatt plant in Texas, with Microsoft and Google building gigawatt-scale plants of their own. Chief executive Peter Gardett argues those buyers are carrying unusual price risk, which would surface either in token prices or on the wider grid.
HD Hyundai Heavy Industries signed a 956 billion won order, worth about $675 million, with the US infrastructure firm Coban Energy Group to supply one gigawatt of on-site generating capacity for American data centers, built around its 9.6-megawatt HiMSEN engines. It is the company's largest power-engine order to date, surpassing a 627 billion won contract it signed in April. The deal is another sign that AI operators are buying their own generation rather than waiting years for a grid connection.
Michael Thomas of the Distilled newsletter identified Amazon as the customer behind GW Ranch, a 35-turbine, 7.65-gigawatt gas plant in Pecos County, West Texas, developed by Pacifico Energy to power a new data center campus. Texas granted the air permit in January, allowing 33 million tons of carbon dioxide a year — more than any existing US power plant, though operators rarely emit their full allowance. Amazon says generating its own power means the site "won't raise electricity costs for Texas families." Land clearing began in late July.
Governor Greg Abbott directed the state's utility commission and its grid operator to verify and audit each data center moving through the interconnection queue, and any project that fails the requirements will be denied a connection. That queue now holds roughly 474 gigawatts of requests, more than five times the state's record peak demand, and about ninety percent of it is data centers. Applicants must disclose projected power draw, on-site generation, water sourcing and cooling method, noise and lighting mitigation, tax incentives received, and who actually owns the project. An earlier voluntary survey went mostly unanswered.
The 69 unpermitted gas turbines powering xAI's Colossus data centers, sited in Mississippi just south of Memphis, will not be fully removed until July 2027 while SpaceX builds a permanent 1.2 gigawatt plant with 41 turbines nearby. Federal rules require permits for turbines that size; SpaceX argues these are exempt because they sit on shipping trailers. The N-A-A-C-P and the Southern Environmental Law Center have sued, pointing to the potential for more than 2,000 tons of smog-forming nitrogen oxides a year in one of the country's more polluted regions.
PJM Interconnection, which serves 67 million customers from Virginia to Illinois, directed staff to build a registry of every site drawing at least 50 megawatts and to curtail those loads first when capacity runs short. New large loads that have not brought their own generation or otherwise secured supply by June 1, 2027, get cut before emergency measures reach ordinary customers, with advance notice ranging from 30 minutes to several days and compensation for participating. Wholesale power prices across the territory have nearly doubled in a year, and PJM now forecasts 70 gigawatts of new large load by 2038.
The Wall Street Journal reports that Nvidia is in talks to guarantee $250 billion in financing so OpenAI can lease a 10-gigawatt campus that SoftBank's energy arm is developing in Piketon, Ohio, on the site of a former uranium enrichment plant. The full project could cost more than $500 billion, and Nvidia is separately discussing another $350 billion tied to chip purchases. Investor Michael Burry, who is short Nvidia, called the structure circular — a chip vendor guaranteeing its own customer's spending on its own chips.
President Trump announced a major expansion of the federal Ratepayer Protection Pledge, adding more than 200 utilities, data-center developers, cooperatives, and states to a commitment the White House says now covers about 80 percent of the power delivered to US homes and businesses. The core promise is that data-center operators — not ordinary customers — pay for the new generation and grid infrastructure their AI campuses require, whether they use that power or not. Duke Energy, among the participating utilities, launched a "Customer Protection Plus" framework; chief executive Harry Sideris told CNBC that a 15-year deal with a single one-gigawatt data center could save its other customers roughly $1 billion — though the utility cautioned that means bills rise less than they otherwise would, not that they fall.
When a power line failed near Washington on July 22, the grid should have recovered in seconds; instead it took more than ten minutes, because roughly 3.1 gigawatts of data-center load — about three percent of the regional grid's demand — disconnected almost at once and swung voltage from Northern Virginia to Chicago. Twice the size of a similar 2024 event, the incident exposed a reliability risk that is the flip side of the affordability fears behind this week's federal ratepayer pledge — two faces of the same data-center power crunch. Proposed fixes include staggered disconnection rules, campus-scale batteries that ride through disturbances, and ride-through mandates already appearing in Texas.
A new BloombergNEF forecast projects US data centers will draw nearly 200 gigawatts of power by 2035 — about 20 percent of the nation's electricity, up from under 6 percent today and roughly four times current demand. The estimate is 83 percent higher than the firm's December projection, driven by a wave of giant AI campuses, more than a quarter of them larger than 500 megawatts. "Every coal plant, every gas plant, every solar farm in the US — one unit of energy out of five generated by them is going to data centers," one BloombergNEF analyst said.
Construction-robotics startup Gritt left stealth with $34 million in total funding, including a $26 million Series A led by Obvious Ventures. Its AI controls off-the-shelf robotic arms and skidders to unload, carry, and place solar panels with sub-millimeter precision. An eight-person crew using the systems installs 3,000 to 4,000 panels a day versus about 800 by hand, and Gritt has contracts to help deploy 2.8 gigawatts of solar over eighteen months. The founders say the robots also cut worker injuries on hazardous outdoor sites.
Governor Kathy Hochul signed an executive order barring new permits for data centers that draw 50 megawatts or more, making New York the first state to pause the AI build-out. The freeze holds until the state finishes an environmental review of energy demand, water use, and noise — a process expected to take about a year. Projects that already hold permits are exempt, and a proposed Grid Acceleration Fund would make operators help pay to upgrade the aging power grid.
A new SemiAnalysis analysis argues the US grid simply cannot add capacity fast enough for AI. Data-center demand is set to climb from 21 gigawatts in 2026 to 84 gigawatts by 2030, while net-new firm capacity is barely 15 gigawatts a year — turning available headroom negative across several regions by 2027. The firm projects that behind-the-meter generation — power plants built on-site, bypassing the grid — will supply well over half of new US data centers from 2028 on, because gas-turbine and interconnection lead times now stretch four to six years and on-site power is both faster and more certain.
Chevron signed a 20-year agreement to supply natural-gas-fired electricity to a massive Microsoft AI data center planned near Pecos in West Texas. The project, called Kilby, would start generating power by 2028 and scale to 2.67 gigawatts — enough for more than 530,000 homes — with a first phase estimated at about $7 billion. Chevron is building the plant with Joulent, an energy firm backed by Engine No. 1, and will make a final investment decision later this year. It is the latest sign that AI's power appetite is reshaping where and how electricity gets built.
In a unanimous vote, the Federal Energy Regulatory Commission ordered the six largest US grid operators to justify or rewrite the rules that govern how AI data centers and other big power users connect to the transmission system. Operators have 30 days to report available capacity and 60 days to defend their regional rates, with data centers footing their own interconnection costs. The orders speed up connections but leave the deeper problem untouched — there still is not enough generating capacity, and data-center demand is on track to nearly triple by 2035.
A Gallup poll released this month found that 70 percent of Americans oppose construction of an AI data center in their local community, including 48 percent who say they are *"strongly opposed."* Half of opponents cite excessive use of resources — 18 percent each name water and energy use specifically — while another 16 percent cite pollution and noise. Two-thirds of supporters cite economic benefits, mostly local job creation. The first national survey of public sentiment on the issue lands as Senator Bernie Sanders has introduced moratorium legislation, Maine's legislature has passed and then watched a governor's veto kill a construction ban, and protests have escalated from Tremonton, Utah, to Monterey Park, California — a sign that the *"AI is good for jobs"* pitch is not yet outweighing local concern about water, grid, and noise.