OpenAI embraces Nvidia and Oracle—will Microsoft be marginalized?

```

As the battle over AI infrastructure enters a new phase, OpenAI has reached unprecedented collaborations with tech giants including Nvidia and Oracle. This has led the market to question Microsoft's future dominance in the GenAI (generative AI) field: Will Microsoft be "marginalized" by OpenAI?

According to ZF Trading Desk, Morgan Stanley stated in a recent report that investor concerns about Microsoft being marginalized in the AI wave are unnecessary. In fact, the bank has upgraded Microsoft stock to "top pick" and raised the target price to $625, representing a 23% upside from the current price.

Analysts believe that Microsoft's selective "letting go" is precisely a sign of its strategic maturity and confidence. In an environment with limited resources such as GPUs and data centers, Microsoft is prioritizing serving enterprise customers with lower customer concentration, larger profit margins, and higher lifetime value.

At the same time, Microsoft and OpenAI have signed a new non-binding memorandum of understanding (MOU), marking that their cooperation is entering the next stage. Morgan Stanley emphasized that Microsoft's growth momentum extends far beyond OpenAI. Its strong Azure cloud growth, solid moat in the enterprise productivity applications sector, and undervalued stock price together offer a very attractive investment prospect.

OpenAI Accelerates Multi-Party Partnerships—How Is Microsoft's Competitive Position Challenged?

Recently, OpenAI signed a series of remarkable infrastructure agreements with several tech giants. The most closely watched is a five-year contract worth up to $30 billion with Oracle, as well as a $10 billion AI computing deployment partnership with Nvidia. In addition, CoreWeave, Google Cloud, and Softbank have also joined OpenAI's list of infrastructure suppliers.

The core of external doubts about OpenAI’s cooperation with other cloud vendors is why Microsoft gave up some large contracts. According to Morgan Stanley, this should be seen as Microsoft maximizing profits in a resource-constrained environment.

The report points out that Microsoft serves a diverse range of enterprise customers, which has multiple advantages compared to binding itself deeply with a single large client (such as OpenAI):

Lower customer concentration risk: Avoiding excessive reliance on a single partner.Higher profit margins: Enterprise customers have more dispersed bargaining power, so there are fewer discounts.Higher customer lifetime value (LTV): Enterprise clients tend to purchase additional software solutions such as databases to build lasting enterprise-class applications, bringing more stable, long-term revenue streams.

Therefore, Microsoft’s choice to shift some of OpenAI’s computing demand to other suppliers may actually reflect the strong demand from its own enterprise customers. Microsoft’s management previously stated that increased capital expenditure is entirely to meet enterprise customer demand. In the most recent quarter, commercial bookings grew 35% year-over-year, corroborating this trend.

Furthermore, though the details of the new Microsoft–OpenAI agreement (such as revenue sharing ratio) have not been announced, the non-binding MOU demonstrates the ongoing evolution of their relationship. Morgan Stanley believes that in the new negotiations, Microsoft will prioritize long-term access to OpenAI's intellectual property and maintain OpenAI’s long-term viability, as a healthy partner is crucial to its own interests.

Diversified Azure Growth: The AI Ecosystem Expansion Far Transcends “Binding” Logic

The sustained rapid growth of Azure’s cloud and AI business is not driven by OpenAI alone. Foundational investments, the wave of enterprise cloud adoption, and GenAI application innovation jointly underpin Microsoft's business resilience.

According to Morgan Stanley's estimates, Microsoft's AI-related capital expenditures have continued to grow in recent years and are expected to do so in the future. Azure AI revenue may reach $10.4 billion in fiscal 2025, and surpass $20 billion in fiscal 2029. The calculations show that even under a scenario of just 30% gross margin, Azure AI's revenue potential far exceeds current conservative estimates.

Morgan Stanley emphasized that AI-related business growth is becoming increasingly diverse: On one hand, Microsoft is signing more and more AI cloud contracts outside of OpenAI, and revenue structure is becoming more distributed. On the other hand, Azure is not only a main cloud supplier for GenAI, but also a comprehensive platform for “enterprise-class” applications, native PaaS and IaaS innovations. CIO surveys show Microsoft maintains a top market share in multiple fields such as enterprise productivity, AI applications, and security. Enterprise users are strongly willing to invest long-term in Azure, M365, and other ecosystems.

Enterprise Applications and Productivity Ecosystem: Microsoft’s Long-Term Moat

Analysts believe Microsoft is not simply a cloud computing provider; its productivity applications (such as M365, Copilot, etc.) and the “digital office + knowledge worker” ecosystem are key to fending off Agentic AI challenges and maintaining high customer stickiness.

Although new intelligent agent–type AIs continue to emerge, Morgan Stanley’s research indicates that CIOs and knowledge workers are showing increasing stickiness toward Microsoft’s M365, Teams, Copilot, and other office applications. Research for the second quarter of 2025 shows that 33% of business customers have already upgraded their main O365 subscriptions to the advanced E5 version, with more than half planning further upgrades; M365 Copilot and other AI products’ organizational penetration is expected to rise from the current 31% to 43%.

Historical experience also shows that Microsoft can always defend its dominant position through mergers/acquisitions and integration (such as Teams’ response to Slack and O365’s response to Google Workspace), as well as deepening its ecosystem—these will also help withstand possible shocks from the “Agentic AI” ecosystem.

~~~~~~~~~~~~~~~~~~~~~~~~

The above selected content is from ZF Trading Desk.

For more detailed analysis including real-time insights and frontline research, please join [ZF Trading Desk Annual Membership]

Risk Disclaimer and TermsThe market has risks, and investments must be made cautiously. This article does not constitute individual investment advice, nor does it take into account the special investment objectives, financial situation, or needs of individual users. Users should consider whether any opinions, viewpoints, or conclusions in this article suit their particular circumstances. Investments made accordingly are at your own risk. ```