OpenAI's Elliptical Strategy: Maximizing Microsoft-Amazon Competition to Gain Dominant Position

On February 27th, OpenAI completed the largest funding round in AI industry history, raising $110 billion with a company valuation reaching $730 billion. This elliptical dual funding— involving the world’s two largest cloud players, Microsoft and Amazon, plus NVIDIA and SoftBank—created an unprecedented dynamic within the AI ecosystem. It’s not just about billions; it’s a strategic positioning game that will shape the industry’s future over the next decade.

The $110 billion fund is divided into four main parts. Amazon leads with a commitment of $50 billion ($15 billion upfront, with the remaining $35 billion to be disbursed gradually upon meeting certain conditions). NVIDIA and SoftBank each inject $30 billion, with NVIDIA planning to return it through the purchase of 5 GW of computing power, while SoftBank provides full cash contribution.

OpenAI founder Sam Altman posted a highly strategic thank-you message on his personal X account: Amazon, Microsoft, NVIDIA, SoftBank. Interestingly, although Microsoft did not contribute in this round, it is mentioned immediately after Amazon—one rank higher than NVIDIA, which invested $30 billion. This choice is deliberate, reflecting the complex relationships among the parties in creating the future AI ecosystem.

Two Phases of Technological Evolution: From Stateless API to Continuous Runtime Environment

Leading AI industry analyst Aakash Gupta identifies that often-overlooked technical details are actually central to OpenAI’s strategy. The focus is on two terms: “Stateless API” and “Stateful Runtime Environment”—each becoming the domain of Microsoft and Amazon within the emerging ecosystem.

Stateless API represents the current AI business model. In this form, each request yields a response; the system does not retain context after the interaction ends. Contemporary applications—smart answer assistants, automatic document summarization, AI-powered search—mostly operate with this model. Its advantages lie in flexibility and ease of integration. Companies in finance, retail, manufacturing, and healthcare can embed AI capabilities into existing infrastructure without major reorganizations. However, as model quality converges and computing costs decline, profit margins from Stateless APIs are increasingly squeezed. Scalability is achieved, but per-transaction profitability continues to decline.

Stateful Runtime Environment, on the other hand, represents the second phase. This environment allows AI agents to retain historical memory, work across long-term tasks, and orchestrate complex workflows. It’s not just about answering questions but acting as a digital workforce executing concrete jobs. Commercial applications of Stateful Environments are still limited, but their potential exceeds feature optimization—they represent a paradigm shift in business models. The scope of investment expands from API call costs to process automation, workflow management, and even a portion of operational labor costs.

Gupta emphasizes that by 2026-2027, nearly all major companies will focus on “autonomous agent workflows” rather than single-use API calls. Those investing heavily in AI will increasingly seek systems that operate continuously, collaborate across platforms, and maintain long-term context. In other words, Stateless API is today’s reality, while Stateful Runtime Environment is the future.

Microsoft Dominates Current Cash Flow

Microsoft announced that its cooperation agreement with OpenAI, set in October 2025, remains in effect. The deal includes OpenAI’s use of Azure services worth $250 billion. Under this new arrangement, Azure remains the exclusive cloud provider for all OpenAI Stateless APIs—regardless of customer or distribution channel, every API call will be processed through Microsoft’s infrastructure. OpenAI’s flagship products, including Frontier, will continue to be hosted on Azure.

This strategy locks in a highly predictable cash flow. Every time users or third-party applications (including those supported by AWS) call OpenAI APIs, billing occurs behind the scenes on Azure infrastructure. Call volume is likely to grow, creating stable revenue growth. However, the challenge lies in the shrinking margin trend—volume growth does not guarantee long-term profitability given competitive pressure on API pricing.

Amazon Bets on Future Productivity

Amazon announced a far more ambitious partnership package. AWS will build a Stateful Runtime Environment with OpenAI, integrated via Amazon Bedrock to enable customers to build and deploy AI agents at scale. AWS will also be the exclusive third-party cloud provider for Frontier, which will run on Amazon infrastructure.

An initial $38 billion agreement between AWS and OpenAI will expand to $100 billion over 8 years. OpenAI will utilize 2 GW of AWS’s Trainium compute power to support Stateful Runtime, Frontier, and other advanced workloads. The collaboration also includes developing custom models tailored for Amazon’s applications.

With a $50 billion cash investment, Amazon isn’t just buying access to technology—they’re securing a position in the AI agent era. As agents become the primary productivity tools in large enterprises, the resources consumed—computing power, storage, scheduling systems, workflow orchestration, cross-tool collaboration—will be concentrated within AWS’s operational environment. This is a bet on the future productivity infrastructure, not just current cash flow.

Elliptical Formation: OpenAI Drives Competitive Dynamics

The combination of these two major cloud strategies creates a highly balanced elliptical shape. One focus on today’s revenue curve (Microsoft), the other on future growth orbit (Amazon). For OpenAI, this is a triumph of distributed betting strategy—not tied exclusively to one cloud provider, not dependent on a single partner’s future, leveraging business prospects to negotiate better terms.

In recent years, OpenAI’s dependence on Microsoft has been significant. Microsoft owns 27% of the company and controls key cloud infrastructure. Early advantages were substantial, but this also gave Microsoft greater bargaining power. Now, with Amazon as a serious competitor, the dynamics shift dramatically.

Neither player can afford to exit this negotiation table. Microsoft cannot allow OpenAI to fully switch to AWS; Amazon cannot let Frontier and future technologies be dominated by Azure. In this elliptical setup—two forces pulling with balanced intensity—bargaining power naturally shifts back into OpenAI’s hands.

Those previously dependent now have constructed a positioning that allows them to continue maximizing resources from both of the world’s largest cloud ecosystems, while maintaining strategic autonomy. This is not just funding—it’s a masterclass in geometric positioning within the hyperscale cloud competition landscape.