Microsoft’s AI Push Beyond OpenAI Could Drive Next Breakout

Key Points

• Microsoft is rolling out proprietary AI models, signaling a strategic shift to lessen dependence on OpenAI.
• In-house LLMs could boost Azure's profitability while accelerating the adoption of Copilot across its ecosystem.
• Deeper vertical integration in AI may solidify Microsoft’s long-term competitive moat.

Microsoft Corporation NASDAQ: MSFT continues to be one of the best-performing companies in the technology sector and the broader market. MSFT stock is up more than 19% in 2025, and over 132% in the last five years. That doesn’t include the company’s rock-solid dividend. All in all, the blue-chip stock has reached the status of a stock you own, not trade.

However, the company does have at least one area of potential concern for investors. A key reason for its growth in the past 18 months stems from its partnership with OpenAI. Earlier this year, the partnership showed its first signs of unwinding. Microsoft AI (MAI), its in-house artificial intelligence (AI) development team, just launched its initial AI models.

• MAI-Voice-1 targets high-fidelity voice generation. Specifically, the model natively generates expressive and natural sounding voice. The model is deployed in several of the company’s consumer-facing features such as Copilot Daily.

• MAI-1-preview is a foundational text model that is designed to help users with everyday queries. It's currently in public testing and the plan is for a gradual rollout for Copilot features.

A Strategic Break from OpenAI

Microsoft and OpenAI have been partners since Microsoft’s initial $1 billion investment in 2019. That initial investment, which has grown to over $13 billion, gave Microsoft access to OpenAI’s generative AI program, better known as ChatGPT.

Microsoft has tightly integrated ChatGPT across its AI stack, using OpenAI’s models to power its Copilot assistants. This partnership gave Microsoft an early lead in the AI race.

But OpenAI has always served as a bridge to Microsoft’s own in-house AI ambitions. The launch of proprietary large language models marks the first step in that transition.

By developing its own LLMs, Microsoft gains greater control over its AI stack—reducing reliance on an external partner, minimizing access risks, and eliminating potential innovation lags tied to OpenAI’s roadmap.

Long-Term Investors May Notice a Familiar Pattern

This appears to be the latest example of a common strategy for Microsoft in which it dominates, then owns critical technology infrastructure. An example of this occurred in the 2010s when Microsoft made its strategic shift to cloud computing with Azure.

To facilitate this transition, the company made several strategic acquisitions that enhanced Azure’s operational efficiency and network capabilities. This vaulted the company to a leadership position in cloud infrastructure.

From there, the company has expanded Azure to support mission-critical workloads for government, financial and industrial clients. Today, Azure has become the foundation of modern enterprise IT.

Do These Models Enhance the Bull Case for MSFT Stock?

Investors may wonder if Microsoft’s pivot to homegrown AI models will deliver a measurable financial impact. One area to watch when the company reports earnings will be its margins in its Azure business. AI workloads are expected to drive revenue growth in this business.

The models also vertically integrate Microsoft’s AI stack, allowing it to better differentiate Copilot and Azure services. This could strengthen the company’s ability to retain existing customers and successfully compete with companies like Alphabet and Meta, which continue to scale their models.

Microsoft is also launching these products at a time when investors are concerned about the company’s valuation. At around 36x earnings, the stock is trading at a slight premium to its historical average.

However, if these models reduce Microsoft’s dependency on OpenAI and accelerate AI adoption across Azure and productivity software, analysts may believe that the premium is justified. That could push the consensus price target above its current level of around $612, which is about 21% above the stock price as of this writing.