OpenAI's $110 Billion Raise: The Largest Private Tech Deal in History—And It Still Might Not Be Enough
Amazon, Nvidia, and SoftBank just poured record capital into ChatGPT's creator. But with $115 billion in projected losses by 2029, OpenAI's cash burn may have already outpaced even this historic fundraise
On Friday, February 27, 2026, OpenAI announced the largest private technology financing in history: a $110 billion funding round that values the company at $730 billion pre-money, or $840 billion fully diluted.
The sheer scale is staggering. For context, the entire US venture capital industry invested $170 billion across all startups in 2023. OpenAI just raised nearly two-thirds of that—alone—in a single round.
Three anchor investors wrote massive checks:
- Amazon: $50 billion
- Nvidia: $30 billion
- SoftBank: $30 billion
And the round remains open. OpenAI expects sovereign wealth funds and other financial investors to add roughly $10 billion more before it closes at the end of March.
CEO Sam Altman appeared on CNBC's Squawk Box hours after the announcement, sitting alongside Amazon CEO Andy Jassy. "We're super excited about this deal," Altman said. "AI is going to happen everywhere. It's transforming the whole economy, and the world needs a lot of collective computing power to meet the demand."
But here's the uncomfortable reality buried in the celebratory press releases: even $110 billion might not be enough.
OpenAI's own projections, reported by CNBC, show the company expects to rack up $115 billion in losses between now and 2029—80% higher than previous estimates. This record-breaking investment will barely cover cash burn for the next three years, let alone fund the compute infrastructure, talent acquisition, and R&D needed to maintain OpenAI's lead as competition intensifies.
The AI race isn't slowing down. It's getting exponentially more expensive. And OpenAI just demonstrated that even being the industry leader doesn't exempt you from the brutal economics of frontier AI development.
The Deal Structure: Who's Paying What, and What They're Getting
This isn't passive capital. Each of the three anchor investors secured strategic partnerships alongside their checks, fundamentally reshaping competitive dynamics in cloud computing and AI infrastructure.
Amazon: $50 Billion for Cloud Dominance
Amazon's investment is the single largest check the company has ever written to any external entity. The structure is phased:
$15 billion upfront, followed by $35 billion when certain undisclosed conditions are met. The Information reported those conditions could include OpenAI achieving AGI (Artificial General Intelligence) or completing an IPO by year-end, though OpenAI hasn't confirmed exact terms.
In exchange, Amazon secured:
Exclusive third-party cloud provider status for OpenAI Frontier: AWS becomes the sole non-Microsoft cloud platform distributing OpenAI's enterprise AI agent platform, Frontier, unveiled earlier in February.
$100 billion AWS expansion: OpenAI will expand its existing $38 billion AWS agreement by an additional $100 billion over eight years—locking in massive, predictable revenue for Amazon's cloud division.
Custom model development: OpenAI will develop customized AI models specifically for Amazon's consumer products, potentially integrating ChatGPT-like capabilities directly into Alexa, Amazon.com, and AWS services.
Trainium chip commitment: OpenAI commits to consuming at least 2 gigawatts of compute capacity powered by Amazon's in-house Trainium AI chips—a strategic win for Amazon's effort to reduce dependence on Nvidia.
Andy Jassy framed the investment as a bet on OpenAI's long-term trajectory. "It's so early right now in the AI space, and OpenAI is off to an amazing start," he told Squawk Box. "They're going to be one of the very big winners, we believe, long term."
Nvidia: $30 Billion for Compute Lock-In
Nvidia's participation resolves months of speculation. In September 2025, reports suggested Nvidia might invest up to $100 billion. By January, those reports shrank to a smaller figure. CEO Jensen Huang pushed back, insisting "we will invest a great deal of money. I believe in OpenAI."
The final commitment: $30 billion, with dedicated infrastructure guarantees.
OpenAI commits to:
- 3 gigawatts of dedicated inference capacity on Nvidia's next-generation Vera Rubin GPU systems
- 2 gigawatts of training capacity on the same architecture
This builds on existing deployments of Hopper and Blackwell systems already running across Microsoft Azure, Oracle Cloud Infrastructure (OCI), and CoreWeave.
For Nvidia, this isn't just an investment—it's a multi-year revenue lock. OpenAI's compute commitments guarantee billions in GPU purchases, cementing Nvidia's dominance in AI infrastructure even as competitors like Amazon (Trainium) and Google (TPUs) try to break its stranglehold.
SoftBank: $30 Billion, Round Two
This is SoftBank chairman Masayoshi Son's second massive bet on OpenAI. He contributed $30 billion to OpenAI's March 2025 round—then the largest private tech deal in history—and just matched that figure again.
Son's conviction in AGI timelines hasn't wavered despite broader market skepticism. With this investment, SoftBank's total stake in OpenAI reaches approximately $64.6 billion, representing roughly 13% ownership in the company.
SoftBank financed the investment through bridge loans and capital raises from major financial institutions, after selling stakes in existing holdings—including, ironically, Nvidia shares—to fund the OpenAI check.
The Microsoft Elephant in the Room
Microsoft, OpenAI's anchor backer since 2019 with more than $13 billion invested, did not participate in this round.
The absence was conspicuous enough that both companies issued a joint statement within hours:
"Nothing about today's announcements in any way changes the terms of the Microsoft and OpenAI relationship that have been previously shared in our joint press releases. Our relationship is as strong and central to both companies as it has ever been."
Microsoft still holds an option to join the round, according to sources familiar with the matter. But the optics are awkward.
Amazon's emergence as a "deep strategic partner"—with AWS gaining exclusive third-party cloud distribution rights for Frontier—creates obvious tension. Microsoft Azure remains the exclusive cloud provider for OpenAI's APIs and first-party products like ChatGPT, but the $100 billion AWS expansion signals a significant strategic shift.
The competitive picture is evolving rapidly:
- Anthropic (OpenAI's main rival) closed a $30 billion Series G on February 12 at a $380 billion valuation—with both Nvidia and Microsoft participating
- Google continues building Gemini internally with massive resources
- Meta is open-sourcing powerful models, undercutting commercial competitors
OpenAI is betting that infrastructure scale, not just model quality, will determine who wins the enterprise AI market. But Microsoft's conspicuous absence raises questions about alignment.
Why OpenAI Needs This Much Money (Spoiler: It's Still Not Enough)
The scale of the raise reflects a brutal economic reality: frontier AI has become incomprehensibly expensive.
The Compute Spending Target
OpenAI has told investors it's now targeting roughly $600 billion in total compute spend by 2030. That's down from an earlier projection of $1.4 trillion, revised after concerns mounted that infrastructure ambitions were outpacing realistic revenue forecasts.
But even $600 billion is staggering. For comparison:
- NASA's entire budget for 2025: ~$25 billion
- Tesla's total revenue in 2024: ~$95 billion
- The GDP of Belgium: ~$600 billion
OpenAI plans to spend as much building AI infrastructure by 2030 as the entire economic output of a G20 nation.
The Cash Burn Problem
Here's where the math gets uncomfortable.
CNBC reported that OpenAI's internal projections show $115 billion in losses between now and 2029—up 80% from previous estimates. That means:
The $110 billion raised will barely cover operating losses for three years, leaving little for the actual compute infrastructure, data center build-outs, GPU purchases, talent acquisition, and R&D required to maintain technological leadership.
Put another way: OpenAI is burning cash faster than it's raising it, even with the largest private tech fundraise in history.
The Revenue Side (The Optimistic Case)
To be fair, OpenAI's revenue is growing explosively.
The company now serves:
- 900 million weekly active users
- 50 million paying consumer subscribers
- 9 million business users
Weekly Codex users (the AI coding assistant) have more than tripled since the start of 2026 to 1.6 million—demonstrating that people are increasingly creating, automating, and shipping software that once required full engineering teams.
OpenAI projects more than $280 billion in total revenue by 2030, split roughly evenly between consumer and enterprise.
If achieved, that would make OpenAI larger than:
- Microsoft's current cloud business (~$100B annually)
- The entire global video game industry (~$200B annually)
- Roughly the size of Walmart (~$650B revenue, but OpenAI's projection is cumulative through 2030)
But there's a massive gap: $600 billion in compute spend versus $280 billion in revenue—a shortfall of $320 billion even in the optimistic scenario.
The Strategic Partnerships: More Than Just Money
The real value of this round isn't the capital—it's the infrastructure commitments and distribution channels.
Amazon's "Stateful Runtime Environment"
As part of the partnership, OpenAI will develop a new "stateful runtime environment" where OpenAI models run natively on Amazon's Bedrock platform.
This is technical jargon for something important: most current AI agents operate through stateless APIs—one prompt, one answer, maybe one tool call, then they forget everything.
Production workflows require:
- Context from previous actions
- Approval chains
- System state persistence
- Error handling
- Safe resumption of long-running tasks
Stateless APIs force developers to build all that orchestration themselves—figuring out how state is stored, how tools are invoked, how errors propagate.
A stateful runtime hands all that infrastructure to developers automatically. It's the difference between building a car from scratch versus buying one that already works.
For Amazon, this means AWS customers can deploy sophisticated AI agents without building custom infrastructure. For OpenAI, it means seamless integration into AWS's massive enterprise ecosystem.
"We have lots of developers and companies eager to run services powered by OpenAI models on AWS," Jassy said. "Our unique collaboration with OpenAI to provide stateful runtime environments will change what's possible for customers building AI apps and agents."
Nvidia's Vera Rubin Commitment
Nvidia's Vera Rubin architecture isn't publicly available yet—it's next-generation beyond the current Blackwell systems.
By securing 5 gigawatts of dedicated Vera Rubin capacity (3GW inference + 2GW training), OpenAI locks in access to cutting-edge hardware before competitors can even order it.
This is the AI equivalent of securing exclusive rights to next-generation fighter jets before they enter production. It creates a performance moat that competitors can't match for years.
The Risks Nobody's Talking About
Despite the record valuation and massive capital infusion, serious risks loom.
1. Conditional Capital
$35 billion of Amazon's $50 billion is contingent on undisclosed milestones. If OpenAI doesn't achieve AGI or IPO by year-end (the rumored conditions), that capital doesn't arrive.
That's a problem if cash burn accelerates faster than expected—which OpenAI's own revised projections suggest is happening.
2. The Revenue-Spending Gap
Even with optimistic $280 billion revenue projections by 2030, OpenAI plans to spend $600 billion on compute. That $320 billion shortfall has to come from somewhere—likely more fundraising rounds, debt, or aggressive cost-cutting that could compromise technological leadership.
3. Antitrust Scrutiny
The FTC previously examined the Microsoft-OpenAI relationship. These new "circular" financing arrangements—where Nvidia invests in OpenAI, which then commits to buying Nvidia GPUs; where Amazon invests in OpenAI, which then commits to $100 billion in AWS spending—are drawing regulatory attention.
Senator Elizabeth Warren has already called for investigations into whether these deals constitute anti-competitive behavior.
4. The Valuation Bubble Question
At $840 billion fully diluted, OpenAI is valued higher than:
- Walmart ($700B)
- Exxon Mobil ($500B)
- Tesla ($800B)
It's approaching the valuations of:
- Amazon ($2.3T)
- Google ($2.1T)
But unlike those companies, OpenAI isn't profitable. It's losing more than $30 billion annually. The valuation assumes:
- Continued technological leadership
- Successful monetization at unprecedented scale
- No major competitive disruptions
- AGI development on schedule
If any of those assumptions fail, the valuation becomes indefensible.
5. Competition Is Accelerating
Google's Gemini 3 launched in November 2025 and has gained significant ground. Anthropic's Claude dominates enterprise coding with specialized tools. Meta's Llama models are open-source and improving rapidly, undercutting commercial pricing.
OpenAI still leads, but the gap is narrowing. And leads in technology markets can evaporate shockingly fast—remember Nokia, BlackBerry, Yahoo.
Conclusion: A Necessary Bet, Not a Sufficient One
The $110 billion raise answers one question definitively: investors still believe OpenAI can win the AI infrastructure race.
Amazon, Nvidia, and SoftBank aren't throwing capital at a science project. They're making calculated bets that OpenAI's technological lead, combined with massive infrastructure scale, will translate into dominant market position and eventual profitability.
But this capital infusion solves the immediate problem—keeping the lights on through 2029—without addressing the fundamental challenge: AI development costs are accelerating faster than revenue growth.
The math is stark:
- $115 billion in projected losses through 2029
- $110 billion raised (plus maybe $10B more)
- $600 billion compute target versus $280 billion revenue projection
Even in the optimistic scenario, OpenAI will need more capital before 2030. Possibly much more.
Sam Altman's vision—AGI that transforms the global economy—requires infrastructure investments at civilizational scale. The $110 billion is just the down payment.
Whether the returns justify the cost remains the defining question of the AI era.
For now, the race continues. And it's getting more expensive by the day.
Sources:
- OpenAI Official Release
- CNBC
- Axios
- TechCrunch
- Reuters
- The Information
- Bloomberg
0 Comments