316
Audio & Video Production295
Software Development223
Automation & Workflow195
Writing & Content Creation178
Marketing & Growth170
AI Infrastructure & MLOps139
Design & Creative146
Photography & Imaging136
Data & Analytics106
Voice & Speech121
Education & Learning117
Customer Support108
Sales & Outreach105
Research & Analysis84
Investors are asking for extra return and protections to buy $14bn of bonds funding a Michigan data center linked to Oracle’s OpenAI deal.
In short: Investors are asking for higher interest and tighter terms to buy $14bn in bonds for an Oracle-backed data center project.
A group of investors has been offered a large bond deal worth about $14bn. The money would help fund a 1 gigawatt data center in Saline Township, Michigan, according to people familiar with the talks. A data center is a building full of computers that runs online services, similar to a giant server room.
The project is linked to a broader agreement in which Oracle would provide OpenAI with large amounts of computing power, which is the raw “horsepower” needed to run AI systems. The Michigan site is being developed by Related Digital, and the bonds are being issued by a special purpose vehicle (a separate company set up just for this project).
Some investors are uneasy because Oracle is a tenant, not a co-owner of the facility. They worry about what happens if construction is delayed or if Oracle ever leaves the lease. To make the risk feel worth it, some investors have asked for yields of more than 1 percentage point above Oracle’s own corporate bonds due in 2040, plus extra protections, including removing a clause that would let the issuer repay the debt early.
Related Digital said it has secured $16bn in financing for the campus, subject to usual closing conditions, and a person familiar with the deal said it could close next week. The sale is being led by Bank of America, and Related Digital said Pimco is the anchor investor. Blackstone is expected to provide about $2bn of equity, according to people familiar with the deal.
AI needs lots of data centers, and they are expensive to build. How these projects get financed helps decide how fast new computer capacity gets added, and how much risk sits with big tech companies versus outside investors.
Source: Financial Times