TLDR
- Microsoft is shifting some Excel and Outlook AI requests to its in-house MAI models.
- The move aims to reduce dependence on external AI providers and lower operating costs.
- MAI models are already processing tens of thousands of Office AI prompts every week.
- Microsoft is expanding its proprietary AI across GitHub Copilot, Teams, and other products.
Microsoft (NASDAQ: MSFT) is taking another step toward reducing its reliance on third-party artificial intelligence providers by routing more AI workloads within its Office ecosystem to its internally developed MAI models.
The strategy is designed to lower the cost of delivering AI-powered features while strengthening Microsoft’s control over the technology that powers its software products.
The shift currently affects certain AI-powered functions in Excel and Outlook, where Microsoft has begun assigning selected user requests to MAI instead of depending as heavily on models supplied by OpenAI and Anthropic. Although the transition remains limited compared with Microsoft’s overall AI infrastructure, it signals a broader effort to develop and deploy proprietary foundation models across its software portfolio.
Lowering AI Operating Costs
According to a person familiar with the matter, Microsoft’s MAI models are already handling tens of thousands of prompts every week across Excel and Outlook. Despite this growing usage, they still represent only a relatively small portion of the company’s total AI processing volume.
The move reflects a practical business objective as Microsoft continues investing billions of dollars in artificial intelligence. Running large language models requires enormous computing resources, making inference costs a significant expense for companies offering AI-powered services at scale.
By increasing the use of internally developed models for suitable workloads, Microsoft can reduce the amount it spends on external AI providers while optimizing performance for specific Office applications.
A Microsoft spokesperson declined to comment on the reported deployment.
Expanding Microsoft’s AI Ecosystem
The Office integration is not Microsoft’s only effort to broaden the reach of its in-house AI technology.
The company has already introduced MAI models into GitHub Copilot, giving developers additional AI capabilities alongside existing model options. At the same time, Microsoft is preparing to launch its own transcription model across Teams and other productivity products in the coming months, further expanding the role of proprietary AI throughout its software ecosystem.
These initiatives suggest Microsoft is pursuing a hybrid AI strategy rather than relying exclusively on outside partners. While collaborations with leading AI developers remain important, Microsoft’s growing investment in MAI indicates it wants greater flexibility over which models power different tasks.
Building specialized models internally also allows Microsoft to optimize AI performance for productivity software while potentially improving efficiency and controlling long-term infrastructure costs.
Less Dependence On Partners
The latest deployment follows comments made in June by Microsoft AI chief Mustafa Suleyman, who said the company was working to reduce its spending on Anthropic.
Although Microsoft continues maintaining relationships with multiple AI providers, diversifying its model portfolio could provide greater bargaining power while reducing dependence on any single external partner.
Anthropic’s Claude models remain available within GitHub Copilot for certain paid subscribers. Higher-tier plans continue offering premium AI capabilities, with some subscription options also featuring different data-retention policies depending on the selected model.
This means Microsoft is not replacing third-party AI entirely. Instead, customers are likely to see a growing mix of internally developed and external models operating behind the scenes, with Microsoft selecting whichever system best matches a particular workload.
Long-Term AI Strategy
Microsoft’s latest AI deployment reflects an increasingly common trend among major technology companies. As generative AI adoption accelerates, firms are looking beyond simply integrating third-party models and are investing heavily in proprietary alternatives that can reduce costs and provide greater strategic independence.
For Microsoft, Office applications represent one of the largest opportunities to deploy these internally built models because millions of users interact with Excel, Outlook, Teams, and other productivity tools every day.
If MAI continues proving capable of handling routine productivity tasks efficiently, Microsoft could gradually expand its role across additional products while preserving access to premium external models where advanced reasoning or specialized capabilities remain advantageous.
The strategy also positions Microsoft to balance innovation with financial discipline, allowing the company to continue expanding AI features without relying exclusively on expensive third-party infrastructure. As enterprise demand for AI-powered productivity tools grows, the success of Microsoft’s MAI rollout could become an increasingly important part of its broader artificial intelligence roadmap.
Get 3 Free Stock Ebooks
Discover top-performing stocks in AI, Crypto, and Technology with expert analysis.
- Top 10 AI Stocks - Leading AI companies
- Top 10 Crypto Stocks - Blockchain leaders
- Top 10 Tech Stocks - Tech giants
