Electricity demand is becoming the clearest measure of the AI industry's extraordinary expansion — and a test of how long that pace can last.

Why it matters: The industry's pursuit of ever-larger models is fueling debate over whether they will deliver enough value to justify mounting environmental and financial costs.

Driving the news: Google, Microsoft and Amazon all highlighted improved energy and water efficiency in sustainability reports released in recent weeks.

• But those gains are being overwhelmed by overall growth.

Stunning stat: Google's electricity consumption rose more than 140% between 2021 and 2025 — already exceeding the outer bounds of projected growth modeled in a 2023 paper by Alex de Vries-Gao, a researcher at VU Amsterdam and founder of online platform Digiconomist.

By the numbers: That scale of growth is across the board.

• De Vries-Gao estimates that the amount of new combined electricity demand added by Google, Microsoft, Amazon and Meta between 2022 and 2025 is roughly twice New York City's annual electricity consumption.

Between the lines: The huge amount of energy required for AI is raising questions about whether the industry's push toward ever-larger models is producing benefits that justify these growing resource demands.

What they're saying: In recent interviews, top sustainability executives at tech companies didn't directly answer if they would reconsider growth in the face of environmental concerns.

• "We are deeply committed to responsibly managing the environmental footprint of our operations," said Kate Brandt, Google's chief sustainability officer.
• "Our goal is not simply to slow the growth of environmental impacts," said Melanie Nakagawa, Microsoft's chief sustainability officer. Instead, the aim is "to reduce the intensity of each unit of growth over time, so that future growth really does become increasingly decoupled from that future impact."

State of play: Much of the AI industry has operated on the assumption that ever-greater scale is going to lead to better performance — and ultimately more profits, said Boris Gamazaychikov, who co-founded Sustainable AI Group, a research and advisory firm that helps companies address the environmental impacts of AI.

The other side: The potential for AI to improve people's lives and curb emissions is playing an increasingly large role in tech companies' sustainability messaging.

Yes, but: Many of the AI applications companies highlight today rely on relatively narrow models rather than the frontier models driving much of today's data-center expansion, according to de Vries-Gao and Gamazaychikov.

• AI companies argue advances in frontier models eventually enable many downstream applications.

Zoom in: Gamazaychikov says AI models should disclose standardized energy-efficiency metrics, similar to fuel economy ratings for cars, so customers can compare how much computing different tasks require.

• "You don't need a Hummer to go to the grocery store," Gamazaychikov said.

Full story

Power giant Constellation Energy's VC arm is investing in Blue Energy, a startup planning to build power plants that will first run on gas before converting to small modular nuclear reactors.

Why it matters: Constellation owns and operates the nation's largest fleet of nuclear plants.

• It's the first investment from its venture arm to advance U.S. small modular reactors, the announcement states.

Zoom out: "It's a huge vote of confidence, more than just the money. It's really a partnership on things like operations and siting development," Blue Energy CEO Jake Jurewicz said in an interview.

Yes, but: The companies are not disclosing the investment size. It comes after Blue Energy closed a $380 million funding round in April led by VXI Capital.

Driving the news: Jurewicz revealed to Axios that Blue Energy has had talks with the Department of Energy's loan office.

• "We are also confident that we uniquely have a path to debt finance, even in the absence of DOE support. But we would readily welcome the DOE support."

State of play: Blue Energy's plan is a repeatable, scalable process — borrowed from the offshore oil and gas and LNG industries — to prefabricate power plants that are then barged to deployment sites in blocks of 1,000 to 2,000 tons.

What's next: It's eyeing 2027 for a final investment decision on its first project in Texas, which would be located adjacent to a Crusoe data center campus.

• It hopes to start providing gas-fired power in 2028 and convert to nuclear by 2031.

How it works: Blue Energy is using GE Vernova gas turbines and GE Vernova Hitachi's BWRX-300 small modular reactor.

• The project model involves assembling key parts of the nuclear plant while deploying gas, enabling conversion to nuclear much faster than typical reactor deployment timelines.

The bottom line: The company argues its model lowers costs and project risks, and provides faster returns for backers.

• "That's a big differentiator. It's a huge 'de-risker' that a lot of the hyperscalers are excited about," Jurewicz tells Axios.

✂️ From the sign of the times files: BP is planning to close its VC arm and divest stakes in over 10 companies to the investment firm Verdane.

• Why it matters: It's the latest sign of the oil and gas giant's pivot back toward its core fossil fuel business, and effort to simplify its structure to improve performance.
• Catch up quick: Past BP Ventures investments spanned EV charging, grid tech, vehicle autonomy and much more.
• What's next: BP will "retain interests in a small number of investments where the technology has the potential to create value for its businesses," yesterday's announcement states.

📢 Michigan Gov. Gretchen Whitmer (D) is pressing data center developers to sign a pledge ensuring costs aren't shifted to ratepayers and vowing that they'll "build, bring, or buy" new generation.

• Why it matters: Whitmer's name is tossed around as a potential White House candidate, though she has recently downplayed the idea. And it illustrates how states are taking different approaches, coming right after N.Y. Gov. Kathy Hochul imposed a one-year moratorium. E&E News has more.

📈 LNG is on track to become the second largest U.S. export industry within five years, trailing only civilian aircraft and parts, the research and consulting firm S&P Global Energy said in a new report.

• The intrigue: Sending more and more gas abroad still has "negligible" effects on U.S. gas prices, the report finds.
• Context: The pro-export U.S. Chamber of Commerce funded the research. But the "analyses, content and conclusions" all come from S&P, it states.
• What we're watching: The Iran crisis is creating even more momentum for investment in LNG projects outside the region, including the U.S. and Canada, S&P analysts say.

AI-related demand within the largest U.S. power grid continues to keep pushing power bills higher for the 67 million Americans served by PJM Interconnection.

By the numbers: The price that generators charge PJM for keeping power ready to supply to the grid at moments of peak demand hit $325 per megawatt-day, according to auction results released Tuesday.

• That amounts to $16.4 billion in costs to be paid by ratepayers.
• Bloomberg notes that the grid's independent market monitor estimates that data centers accounted for roughly $6.3 billion of that total.

Full story

That's China's share of rare earth element refining last year, falling from 90% in 2024, with the U.S. claiming the bulk of its lost share, per a new International Energy Agency report.

Why it matters: China's dominance gives it powerful economic and geopolitical leverage over the materials used in key defense and energy applications.

What's next: China's share could keep slipping significantly, falling to 73% in 2035 in IEA's base case projection.

Yes, but: Geographic concentration in the wider critical minerals space is increasing overall.

Full report