New Jersey fines an AI data center $1.07M over 62 gas generators

Share
New Jersey fines an AI data center $1.07M over 62 gas generators

The enforcement bill for the "bring your own power" data center buildout just arrived in New Jersey — and it came with a number small enough to be telling. Meanwhile, Tokyo's Sakana AI put the man who invented the LSTM on its letterhead.


New Jersey's Department of Environmental Protection has fined DataOne $1.07 million for installing 62 natural-gas generators at its Vineland site without a preconstruction permit, and for running them without a valid operating certificate. The penalty order says the state has never levied a larger one against a data center: "by far the largest ever taken against a data center in New Jersey and possibly one of the largest such actions in the nation," DEP commissioner Ed Potosnak said in the announcement.

The hardware is what makes the case unusual. The order lists 62 Caterpillar G3516H natural-gas engines rated at 1,982 kilowatts each — roughly 123 megawatts of on-site generation, where New Jersey requires a permit for anything from 37 kilowatts up. The facility at 3963 South Lincoln Avenue in Cumberland County is being developed for the Nebius Group to supply Microsoft's AI compute under a $17 billion deal, and is planned to reach 300 megawatts behind two schools and a residential neighborhood about 40 miles outside Philadelphia. DataOne says it is moving to Bloom Energy fuel cells as the permanent power source and disputes that the gensets were a permanent installation — "while we disagree with the temporary generator determination, we will apply for the air permits," a spokesperson said. It has 45 days to file or stop running them.

Two details are worth keeping straight, because they are already circulating wrong. The evidence came from thermal drone footage — Floodlight reporter Evan Simon, published with The Guardian on August 27 — showing at least 45 of the 62 engines running, not the satellites that some headlines imply; the satellite enters only as the original tip, from a resident who noticed hardware imagery while looking at a nearby golf course. And the trigger was DEP's own July 29 inspection, not the news coverage. Virginia fined a Microsoft data center $2.5 million the week before, and this is the first penalty aimed squarely at the trend now touching roughly 60 US data centers: building the plant on site and asking about the air permit afterwards. A million dollars against a $17 billion compute contract is a fraction of a rounding error, which is exactly the point critics are making.


Sakana AI has hired Jürgen Schmidhuber as chief scientific advisor, where he will help steer the Tokyo lab's RSI research group — the team Sakana says is meant to "redesign the AI development process itself with AI." Schmidhuber's 1991 work underpins much of modern deep learning, and Sakana credits his 1990 world-model research and 1987 thesis on recursive self-improvement as the direct ancestors of its Darwin Gödel Machine and AI Scientist projects. The framing to note: he joins alongside his existing posts and will travel to Tokyo regularly, so this is a marquee name on a lab announced in June — not a new capability milestone. The more interesting signal is Sakana's stated constraint, that its advances should compound on national rather than hyperscale compute budgets, which is a different bet than anything coming out of California.

What to watch: whether other states copy New Jersey's enforcement template, and whether DataOne's 45-day clock ends in permits or in shutdown.

Is a seven-figure fine against a $17 billion AI buildout a real deterrent, or just the cost of doing business? Tell us in the comments.

Read more

Akhetonics says its all-optical CPU reaches a customer in 2026

Akhetonics says its all-optical CPU reaches a customer in 2026

Light-based computing keeps promising more than it delivers — but one Munich startup has just put a date on its bet, and the interview laying it out is doing the rounds on Hacker News this week. Akhetonics says it will deploy its first commercial machine with a major customer by the end of 2026, with several more planned for 2027. The company, founded by Michael Kissner and Leonardo Del Bino, is building a computer where data enters as light, is switched as light, and circulates through memory

The Week in AI — October 5–11, 2026

The Week in AI — October 5–11, 2026

Every big claim this week turned out to rest on fine print more interesting than the headline: revenue only the company reporting it can define, safety tests sandboxed while the product keeps the web, and a Pentagon phase-out nobody would confirm until reporters kept asking. The week's top 5 1. OpenAI's revenue was $20 billion below the numbers everyone quoted — and the gap was definitional. The Financial Times reported Thursday that OpenAI's annualized revenue runs roughly $20 billion unde

Drone strike shuts a third Yandex data center, taking YandexGPT offline

Drone strike shuts a third Yandex data center, taking YandexGPT offline

Russia's largest tech company is learning what the AI era's infrastructure war looks like from the receiving end — three data centers in four days, and with them much of the cloud layer Russian businesses run on. A Ukrainian drone strike knocked out Yandex's data center in Vladimir early Sunday morning, the third of the company's facilities hit since October 8. The site — reported at roughly 50 MW and designed for about 2,880 server racks — stopped operating completely after the attack, Yandex

Agent teams cost up to 5x more, barely score higher

Agent teams cost up to 5x more, barely score higher

The multi-agent hype train hit a benchmark this weekend — and the grid and the trucking regulators had quiet weeks of their own. Vals AI put agent teams head-to-head with single agents on its Vibe Code Bench, and the teams cost between 1.8 and 5.1 times more for almost no extra quality. The evals company ran GPT-6 Sol and Claude Opus 5.5 solo and in teams across 50 apps at two reasoning efforts; out of four comparisons, only one was statistically significant — Sol at medium effort, where the t