$110 Billion Financing is Essentially a Gamble: OpenAI Must Achieve AGI or IPO This Year
The deal and the deadline
OpenAI closed an unprecedented $110 billion financing round at the end of February 2026 — reportedly the largest single private financing in history. The round left OpenAI valued at about $730 billion pre-money and roughly $840 billion post-money, figures that eclipse many public stalwarts (JPMorgan’s market cap is around $830 billion; SpaceX has been valued near $800 billion). It has been reported that Amazon’s roughly $350 billion tranche is conditional: the funds would only be released if OpenAI achieves artificial general intelligence (AGI) or completes an IPO by the end of the year. Who’s really buying shares, and who’s buying leverage?
Burn, market share and why investors don’t care about the losses
The numbers are stark. OpenAI’s revenue in 2025 reportedly reached $13.1 billion, beating a $10 billion target. Yet losses are ballooning — about $8 billion in 2025 and a projected $25 billion in 2026, with internal forecasts pointing to cumulative losses into the hundreds of billions through 2029 and profitability pushed into the 2030s. ChatGPT’s market share has also slipped, from roughly 86.7% in early 2025 to an estimated 64.5% by early 2026 as rivals such as Anthropic and Google gain ground. So why would Amazon, NVIDIA and others pour in cash? Because much of this financing reportedly returns to the investors as services and hardware — GPUs from NVIDIA, cloud infrastructure from Amazon and Microsoft, and financing or facilities from SoftBank — effectively creating a capital-and-resource closed loop that secures compute and influence rather than a pure equity bet.
An infrastructure race with geopolitical undertones
This is less a conventional startup round and more a bid for compute domination. OpenAI has reportedly told investors it will spend roughly $600 billion on compute by 2030, and CEO Sam Altman has floated even larger infrastructure ambitions. The White House-backed “Stargate” plan (星门计划), which promised joint investments in AI data centers, has been reported stalled by fights over control between OpenAI, SoftBank and Oracle — a reminder that ownership of physical data centers matters as much as model IP. Add U.S. chip export controls and rising tech-mistrust between Washington and Beijing, and the stakes become geopolitical: who controls OpenAI may shape future standards, commercial rules and everyday AI experiences. In short: investors are not just buying a company — they are buying an option on the next-era infrastructure, and they want that option to convert into either AGI or a public exit this year.
