The Trillion-Dollar Question: OpenAI in Early Talks for a Funding Round Above $1.2 Trillion

Less than six months after closing the largest private funding round in history, OpenAI is reportedly already laying the groundwork for an even bigger one. Bloomberg reports that the ChatGPT maker is holding early talks with investors about a fresh funding round that would value the company at more than $1.2 trillion — a figure that, if even approximately achieved, would make OpenAI the most valuable private company on Earth by an enormous margin, and the first to cross the trillion-dollar line in private markets.

The Financial Times and The Information corroborated the story within hours, adding a detail that matters more than the headline number: according to the FT, the talks were initiated by investors rather than by the company. OpenAI is not out banging the tin cup — its bankers and executives are fielding inbound demand from institutions that fear being locked out of the last private round before an IPO.

What we actually know

The reporting, published September 15–16, 2026, is consistent across three outlets on the core facts:

• The talks are early. Nobody has signed term sheets. Structuring, sizing, and pricing are all unresolved, and the round could still change shape materially or dissolve.
• The valuation discussed is “more than $1.2 trillion.” That is roughly a 40% step up from the $852 billion post-money valuation OpenAI locked in on March 31, 2026, when it closed $122 billion in committed capital led by SoftBank and Microsoft.
• Investors initiated the discussions. The FT’s framing suggests scarcity dynamics — investors who missed the March round, or who want to increase exposure before a listing, are driving the approach.
• The round is explicitly framed as pre-IPO positioning. It would set a reference price for the public debut that Altman has already said will not happen in 2026.

That last point is the connective tissue of the story. On September 12, Sam Altman told Fortune that an IPO “now would come at an ill-advised moment,” citing AI safety concerns, and confirmed the company would not go public this year — while his CFO, Sarah Friar, has publicly targeted a 2027 listing, with a confidential filing potentially as soon as late 2026. A private round at $1.2 trillion+ is precisely how you bridge that gap: you let private capital mark the company up on your own timetable, then arrive at the IPO with the price already established.

The arithmetic of a trillion dollars

To understand how audacious the number is, run it against the financials that have leaked into public view. OpenAI’s annualized revenue is estimated at roughly $25 billion as of mid-2026, against substantial losses — analysts project the company will not be profitable until around 2030, with some estimates putting cumulative losses in the hundreds of billions by then as compute buildouts compound. A $1.2 trillion valuation on $25 billion of revenue is a ~48x revenue multiple for a company that is deeply cash-flow negative.

For context, at the top of the 2021 software bubble, elite high-growth SaaS companies traded at 40–50x forward revenue — while being profitable or close to it. OpenAI’s supporters would counter that no software company ever grew at OpenAI’s slope, that its revenue run-rate reportedly tripled year over year, and that “AI operating system” economics are not comparable to seat-based SaaS. Its critics — and there are many, including some of its own investors — argue that the entire private valuation stack rests on circular financing: chipmakers, hyperscalers, and sovereign funds investing in the labs that then spend the money buying chips and compute from those same partners.

The $122 billion March round already embedded this tension. One investor in both OpenAI and Anthropic noted at the time that “the only way OpenAI’s last round makes sense is if it goes public above $1.2 trillion.” This week’s talks are, in a sense, that logic executing itself: the private market is now marking the company at exactly the number required to make the previous round’s math work.

Why investors are lining up anyway

Three structural forces explain the inbound demand:

Scarcity. OpenAI equity is one of the few assets offering pure-play exposure to frontier AI at scale. The March round was heavily allocated to strategics — SoftBank, Microsoft, Middle Eastern sovereign wealth funds — leaving generalist institutions underweight. Secondary trades have priced the stock around $908 billion in recent months, and Anthropic’s secondaries reportedly jumped to a $1.2 trillion equivalent valuation in July, briefly overtaking its rival. Nobody wants to be the fund that explains to LPs why it passed twice.

The pre-IPO window. Once OpenAI files publicly, the discount disappears. A final private round is the last chance to buy below the IPO reference price — or at least, below the range the bankers will point to. Investors initiating talks now are effectively bidding for a place in the comparator set.

The infrastructure story. OpenAI has committed to roughly $1.2 trillion in cumulative infrastructure spending across Stargate and related ventures through the end of the decade. That buildout requires capital far beyond what the March round provides, and equity is the most expensive but most durable way to fund it. Investors buying in at $1.2 trillion are, knowingly or not, pre-funding concrete and gigawatts.

What could break it

Early talks collapse more often than they mature. The obvious failure modes: a macro shock that closes the window for mega-rounds; a disappointing GPT-cycle slowdown that makes the revenue slope look mortal; or internal disagreement at OpenAI about whether to take the markup at all — a $1.2 trillion private price is also a $1.2 trillion expectation the IPO must clear, and an IPO priced below the last private round would be a humiliation that Altman, watching what down-rounds did to other pre-IPO darlings, is keen to avoid. His public caution about timing (“ill-advised moment”) suggests he knows the mark cuts both ways.

There is also the Anthropic factor. If the race at the frontier stays as close as 2026 has shown it to be, being the first to $1.2 trillion is a psychological prize but not a decisive one — and the market may eventually price the two labs as a duopoly rather than a winner-take-all, compressing both premia.

The bottom line

None of the money has moved yet; these are talks, not terms. But the direction is unmistakable. OpenAI’s private valuation went from $157 billion in late 2024 to $300 billion in early 2025, $500 billion by that autumn, $852 billion in March — and now investors themselves are proposing $1.2 trillion. Each round has been led by demand rather than necessity, and each has been absorbed. Whatever your view of the fundamentals, the machine that converts narrative into capital at trillion-dollar scale is operating at full throttle, and the only remaining question is whether the public markets, in 2027, will agree to keep it running.