Microsoft is charting a bold new course in the artificial intelligence race, and CEO Satya Nadella is speaking openly about how the company plans to dominate the space on its own terms. In a frank conversation tied to Microsoft Build 2026, Nadella acknowledged that the tech giant leaned too heavily on its partnership with OpenAI in the early days of the AI boom — and that this reliance ultimately slowed its competitive edge.
Now, Microsoft is fighting back with a powerful lineup of homegrown AI models under its MAI brand. The company revealed seven new in-house models at Build, including MAI-Thinking-1, a cutting-edge reasoning model, and MAI-Code-1-Flash, its first coding-specific AI model. These developments signal a major shift in how Microsoft approaches artificial intelligence — no longer as a consumer of others’ technology, but as a creator in its own right.
Nadella described what he called a “dual frontier” strategy, where Microsoft maintains its own model lineage while continuing to collaborate with OpenAI through 2032. The company is also being strategic about its computing resources, prioritizing enterprise clients and internal model development over selling raw GPU power to smaller AI startups.
The strategy seems to be paying off. Microsoft reports that nearly 90 percent of Fortune 500 companies now use AI agents built with its tools, and the company’s AI business has crossed a remarkable $37 billion annual revenue run rate.
With its sights set on coding, cybersecurity, and knowledge work as its three core AI domains, Microsoft is positioning itself not just as a participant in the AI revolution — but as one of its defining architects.
