Oracle's $7 Billion Wisconsin Power Guarantee Isn't Really About Power
Oracle's power guarantee for a nearly one-gigawatt Wisconsin AI campus was never really a story about electricity. Oracle says it may ultimately have to post more than $7 billion in financial security, likely in the form of a letter of credit, at a financing cost that could exceed $100 million a year, to back power commitments behind the Lighthouse Campus it's building in Port Washington with…
Oracle's $7 billion Wisconsin power guarantee for a one-gigawatt AI campus isn't about electricity. The financial security, likely a letter of credit, could exceed $100 million annually due to power commitments backing the Lighthouse Campus being built in Port Washington with Vantage and OpenAI. Oracle requested the commission to reconsider, but the commission declined.
The dispute isn't about power generation or Oracle's balance sheet; it's about who bears the consequences when infrastructure commitments are made without verifying future demand. Every large infrastructure commitment starts as a promise about future consumption, made by a party not paying the full cost of being wrong. Planning systems have to act on these promises before the consumption exists, leading to a failure where infrastructure decisions can't wait for realization.
Wisconsin's regulators tightened rules for data center developers, requiring collateral sized to the power infrastructure built to serve the customer. Oracle's S&P rating was downgraded to BBB-, below the exemption bar. The regulator didn't dispute the campus will likely be built; instead, they refused to let the possibility of it not materializing sit with ordinary electricity customers.
Oracle became the forced arbiter, absorbing the financial consequences of a forecast that didn't convert into consumed power. This pattern, called Phantom Capacity, shows how capacity commitments are becoming financial instruments and how the underlying validation problem remains unsolved.
Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.