OpenAI has closed on roughly $11 billion in new funding, with the money already wired to the company at a $1.25 trillion valuation — and the round, initially pitched five weeks ago at $30 billion or more at $1.4 trillion, may top out at just $20 billion. Bloomberg broke the news Friday, reporting the funds landed as of Friday local time, with the remainder of the round expected to close in the coming weeks.
🔍 THE BOTTOM LINE
The largest AI fundraise in history is landing smaller and cheaper than its target: $11 billion banked at $1.25 trillion, versus $30 billion sought at $1.4 trillion in September. The valuation itself is still astonishing — nearly 50% above the $852 billion OpenAI raised at in March — but the delta is the story. The world’s most sought-after private AI company just took a visible haircut from the same market that is now repricing Anthropic’s IPO weeks before listing.
What actually happened
Per Bloomberg’s reporting, carried in full by Yahoo Finance, people familiar with the matter said the ~$11 billion was wired to OpenAI as of Friday. The initial target when the round kicked off in late September was “$30 billion or more” at a $1.4 trillion valuation. The reduced goals, Bloomberg notes, reflect “a stark shift for a fundraising juggernaut that helped usher in the current AI boom” — with sentiment souring over infrastructure spending and the prospect of new roadblocks to data-centre construction after the US midterms.
Two reference points make the scale legible. OpenAI’s March round raised $122 billion at an $852 billion valuation — so the new money is less than a tenth of that earlier haul. And as TechCrunch’s Lucas Ropek reported, the round “hasn’t closed yet” and could total about $20 billion — well short of the $30 billion Bloomberg reported OpenAI was chasing in late September.
the-decoder’s summary of the WSJ reporting last week identified where the gap came from: fading interest from major institutional investors amid the tech selloff that followed weak quarterly results from Google, Amazon and Intel, with OpenAI planning to fill the remainder from family offices and high-net-worth individuals — a different, shallower buyer pool than the sovereign funds and crossover giants that powered the March round.
OpenAI declined to comment to Bloomberg. The company had previously signalled it would go public this year, then delayed; expectations now sit on an IPO sometime next year.
The same repricing is hitting Anthropic’s IPO
The timing is not a coincidence, and the juxtaposition landed within hours. Buyers are repricing in public. CNBC reported Friday that some Anthropic investors now expect the listing below $1.5 trillion, with some warning it could dip to $1 trillion — against the roughly $2 trillion target when the company filed confidentially with the SEC. Bloomberg’s separate reporting put buyer estimates as low as $1.1 trillion. Whichever floor you pick, the direction is the same: the first time the market gets to price a frontier lab against public-market scepticism rather than private-round enthusiasm.
The two labs now sit in a squeeze from both directions. OpenAI banked paper valuation at $1.25 trillion private; Anthropic faces a public listing that buyers want to enter at $1.1 trillion or below. Whichever way the IPO prices, it becomes the reference point for the entire private AI market — including the OpenAI round that just closed.
Anthropic’s Q3 numbers, disclosed in its amended S-1 earlier this week (covered here), frame what buyers are being asked to believe: nearly $17 billion of quarterly revenue, $23 billion of quarterly losses, and a compute-contract load measured in the hundreds of billions. The rout that began after the midterms — where voters rebuked data-centre construction in 40 of 41 ballot measures (covered here) — is the backdrop OpenAI’s own CFO addressed directly on Friday: capital raising continues regardless of how communities vote (covered here).
What the haircut actually tells you
Strip the framing and the arithmetic is blunt. In September, $1 billion of OpenAI equity was priced at roughly $1.4 trillion of company. In November, the same billion buys $1.25 trillion. That is a ~10.7% markdown on the world’s hottest private asset in five weeks — not a collapse, but the first time OpenAI’s price has moved down between rounds since the AI boom began, after March’s $852 billion and the $122 billion investors pushed in at that round.
The round’s composition shift matters as much as the size. Bloomberg reports the remainder is expected to come from family offices and high-net-worth individuals — smaller cheques, thinner diligence, faster wires. When the deepest-pocketed institutions step back from the company defining the boom, and the round is completed by private wealth looking for access, that is a pattern visible in every prior bubble’s late stage. It does not prove the top is in. It does mean the marginal buyer has changed.
The NZ and reader angle
For readers tracking the singularity ledger rather than the market tape, the useful signal is this: the labs’ capability narratives and their funding reality have decoupled. This same week, an unreleased Anthropic model was announced as having cracked an 18-year-old quantum computing conjecture (covered here), and METR measured Anthropic’s research accelerating at 1.9x historical rate with 80% of the gain AI-attributed (covered here). The machines keep getting better on schedule. The money is arriving later, smaller, and from less institutional hands. Both can be true for a while — history says not forever in the same direction.
The AI trade now hangs on one near-term event: Anthropic’s IPO pricing, expected within weeks. If it lands near the $2 trillion buyers were quoted in September, the OpenAI round looks like bargain pricing. If it lands near $1.1 trillion, OpenAI’s private mark is suddenly the optimistic one — and the entire 2026 class of AI mega-deals reprices around it.
As of publication, OpenAI has not commented on the round; figures come from Bloomberg’s unnamed sources familiar with the matter, and the final round size is not yet locked.
Sources: Bloomberg via Yahoo Finance · TechCrunch · CNBC · the-decoder
