OpenAI at $852 Billion: The World's Most Expensive Bet on AI

On the last day of March 2026, OpenAI announced it had closed a funding round of $122 billion at a post-money valuation of $852 billion — the largest private fundraise in corporate history. The figure is so large it strains comprehension. For context, it places OpenAI ahead of Walmart, Visa, and JPMorgan Chase by market capitalisation, despite the company never having turned a profit.

The round was anchored by a trio of strategic giants: Amazon contributed $50 billion, Nvidia $30 billion, and SoftBank $30 billion. Andreessen Horowitz, D.E. Shaw Ventures, Abu Dhabi's MGX sovereign wealth fund, TPG, T. Rowe Price, and Microsoft also participated. In an unusual move, OpenAI extended access to individual investors through bank channels for the first time, raising an additional $3 billion from retail participants — a clear signal that the company is building its public market narrative ahead of an anticipated IPO later this year.

"We are entering a new phase where frontier AI moves from research into daily use at global scale. Leadership will be defined by who can scale infrastructure fast enough to meet demand." — OpenAI, April 2026

The Numbers Behind the Headline

Strip away the spectacle and the underlying financials tell a more complicated story. OpenAI is generating $2.6 billion in revenue per month — impressive by almost any measure — but the company is burning through cash at a rate that keeps profitability firmly out of reach. Internal projections suggest the company will burn approximately $9 billion in 2025 and $17 billion in 2026, with cash-flow positivity not expected until 2030 at the earliest.

Revenue has accelerated sharply nonetheless. OpenAI made $13.1 billion in 2025, and the company claims it is growing revenue four times faster than Alphabet and Meta did at comparable stages. ChatGPT now counts more than 900 million weekly active users and over 50 million paying subscribers. Enterprise business — API access, corporate ChatGPT plans, and AI deployment for large organisations — now accounts for 40 per cent of total revenue, up from around 30 per cent a year ago, and is on track to reach parity with the consumer side by the end of this year.

An advertising pilot launched quietly earlier this year has already generated more than $100 million in annualised recurring revenue within its first six weeks, opening a revenue stream that could transform the company's economics if scaled. It is a striking development for a company that built its reputation — and its user base — entirely without advertising.

The Road to an IPO

The April fundraise is as much about building an IPO story as it is about the capital itself. OpenAI has hired Cynthia Gaylor, former CFO of DocuSign, as its first head of investor relations, and internal targets reportedly include filing in the second half of 2026 with a listing in 2027 at a valuation of up to $1 trillion. The decision to allow retail investors into this private round via bank channels — and the company's planned inclusion in several ARK Invest ETFs — is a deliberate effort to broaden the shareholder base before it hits public markets.

The restructuring of OpenAI's corporate form is also worth noting. In October 2025, the company completed a recapitalisation that converted its for-profit subsidiary into OpenAI Group PBC — a public benefit corporation — while keeping the renamed OpenAI Foundation nonprofit in ultimate control. The Foundation's equity stake in the new structure was valued at approximately $130 billion at the time. Microsoft's stake, representing roughly 27 per cent on a diluted basis, was valued at around $135 billion.

The Energy Problem That Won't Go Away

Whatever the headline valuation, OpenAI's most urgent operational challenge is not revenue growth or governance structure. It is power.

On 9 April 2026 — just ten days after closing its record-breaking funding round — OpenAI quietly announced it was pausing Stargate UK, its flagship British infrastructure initiative. The project, announced with considerable fanfare in September 2025 alongside Nvidia and UK cloud provider Nscale, had planned to deploy up to 8,000 high-performance GPUs across sites in north-east England, including the government-designated AI Growth Zone at Cobalt Park. The longer-term ambition was to scale to 31,000 GPUs and provide local computing power for the NHS, regulated financial institutions, and national security applications.

The reason for the pause was blunt: energy costs. UK industrial electricity prices are among the highest of any developed economy — more than four times those in the United States, Norway, Finland, or Sweden. For a data centre drawing 100 megawatts of continuous power, that differential is not a line item. It is a structural problem that fundamentally changes the investment case. Grid connection timelines compound the issue: connection requests in the UK surged from 41 gigawatts in late 2024 to 125 gigawatts by mid-2025, with an estimated 75 gigawatts attributable to data centre projects. Buildings can be completed in 18 to 24 months; grid connections routinely take three to eight years.

A second obstacle is the UK's unresolved approach to AI copyright. The government's preferred mechanism — a broad text and data mining exception with an opt-out for rights holders — was roundly rejected by the creative industries in early 2026. With the policy now delayed until at least 2027, the legal environment for training AI models on UK-sourced content remains unclear. OpenAI cited both factors in its statement: the company said it would move forward "when the right conditions such as regulation and the cost of energy enable long-term infrastructure investment."

The UK government, which had championed Stargate UK as a centrepiece of its AI strategy and cited it as part of a broader £150 billion technology investment package, described the decision as disappointing. For Nscale, the British startup that had staked considerable reputational capital on the partnership, the pause represents a significant setback.

Where the Money Is Going Instead

The contrast with Stargate Norway is instructive. OpenAI's Nordic project — announced in July 2025 in partnership with Nscale and Aker — is progressing. Located outside Narvik in northern Norway, the site benefits from abundant, inexpensive hydroelectric power and a stable regulatory environment. The ambition is to host 100,000 Nvidia GPUs and create what the company is billing as Europe's first AI gigafactory. That, not Cobalt Park, appears to be where European AI infrastructure investment is flowing.

In the United States, the original Stargate project remains on track. OpenAI has deepened its compute relationships across multiple cloud and chip partners: Microsoft Azure, Oracle, AWS, CoreWeave, and Google Cloud on the infrastructure side; Nvidia, AMD, AWS Trainium, Cerebras, and a custom chip in partnership with Broadcom on silicon. The company describes this multi-vendor, multi-cloud approach as a deliberate strategy to avoid concentration risk and maintain flexibility as model architectures evolve.

The Harder Question

The $852 billion valuation has attracted the scepticism it invites. OpenAI is priced at a multiple of its revenue that traditional finance frameworks struggle to accommodate. It will not be cash-flow positive for at least four years. It is scaling infrastructure in a world where the cost of energy is becoming a hard constraint on AI ambition. And it is preparing to face public market scrutiny — with all the quarterly reporting and analyst pressure that entails — at a moment when the gap between its stated potential and its current financials is still very wide.

None of that makes the bet obviously wrong. The company did reach 100 million users faster than any platform in history. Its enterprise revenue is growing quickly. GPT-5.4, its latest model, is driving accelerating adoption in agentic workflows. And the investors backing it — Amazon, Nvidia, SoftBank, Microsoft — are not known for writing $30 billion cheques on sentiment alone.

But the Stargate UK pause is a useful reminder that the bottleneck in the AI race is no longer ideas or capital. It is kilowatts. The country — and the company — that solves the energy problem at scale will have a structural advantage that no funding round can replicate.

OpenAI has the money. The question is whether it can find the power to spend it.