The News
The Office Of Strategic Capital Weighs Its Largest Loan Ever
On September 10, 2026, the Wall Street Journal reported that the Pentagon is in talks to lend roughly $5 billion to the AI cloud-computing startup Fluidstack, adding, in its words, the U.S. military to the growing list of financiers of artificial-intelligence infrastructure. Reuters syndicated the report while noting it could not immediately verify it.
The money would come from the Pentagon's Office of Strategic Capital, which lends to companies in areas deemed critical to national security. A loan of that size would be by far the office's largest to date.
What it funds is the notable part. Fluidstack would use the loan to strengthen U.S. supply chains and manufacturing capacity for certain data-center components rather than to build a new AI facility. In practice that means the electrical apparatus sitting between a signed compute contract and a delivered megawatt.
Neither the Defense Department nor Fluidstack commented, and nothing has been signed.
Forget SpaceX. Elon's next BIG bet is a radical "light-speed" device that's turning AI into "Accelerated AI" – and making it 100 times faster and 100 times more energy efficient.
The mainstream hasn't caught on yet. But "Accelerated AI" stocks are breaking out as we speak: 133%, 217%, or even 320% or more… and it's just getting started.
The Company Behind It
A Neocloud That Grew Faster Than Its Own Valuations Could Track
Fluidstack was founded in 2017 at Oxford University and moved its headquarters to New York in late 2025. It builds and operates large-scale GPU clusters for AI companies rather than selling general-purpose cloud services.
Its valuation has moved sharply. The company raised an $830 million round at a $7.5 billion valuation earlier this year, and by September the Journal reported a $1.5 billion round led by Jane Street valuing it near $18 billion. A $5 billion loan would equal roughly 28% of that mark.
Its customer list explains the interest. Fluidstack has worked with Mistral, Character.AI, Poolside, and Black Forest Labs, and in late 2025 Anthropic announced a $50 billion plan to build custom U.S. data centers with the company, starting in Texas and New York.
Why This Matters Financially
Government Money Changing The Cost Of Capital
The loan would lower the cost of building components that private lenders price as cyclical industrial risk. When Washington decides an industry is strategic, projects no longer need to clear purely private hurdle rates.
The target is deliberate. Compute contracts are worthless without transformers, switchgear, and cooling equipment, and those have become the binding constraint, so subsidizing the components unlocks far more capacity than funding any single facility.
There is a structural reason for that framing too. The Office of Strategic Capital's eligible technology list covers advanced manufacturing and microelectronics but not data centers, so financing the supply chain is the available route.
Limits and Uncertainty
The Catch: Nothing Is Signed, And Cheap Capital Cuts Both Ways
This is a report of ongoing talks. Terms, size, and structure could change materially or the discussions could collapse, and both parties have stayed silent.
The deeper question is what government lending does to discipline. More than $400 billion in bonds has already been raised for data centers and AI projects this year, so this is not rescue financing for a sector private markets have abandoned. Public money on top of that can make one borrower safer while making the whole industry harder to underwrite, because it lets construction run ahead of economics that have never been tested. Fluidstack itself is a young company whose valuation more than doubled this year, taking on an obligation worth a large fraction of its worth.
The talks matter because they show AI infrastructure crossing from a private spending race into national industrial policy. The real impact depends on whether a loan is signed at all, and whether subsidized capital fixes a genuine bottleneck or simply accelerates a buildout whose demand has not yet been verified.
Disclosure: This content is for educational and informational purposes only and does not constitute investment advice or recommendations. You should always conduct your own research or consult a qualified financial advisor before making investment decisions.
