Satya Nadella transformed Microsoft into a cloud and subscription leader after becoming CEO in 2014. Now, the company faces a tougher challenge in the AI era.
Despite early investment in OpenAI and a large Azure AI business, Microsoft has yet to produce a breakout AI model or application on par with OpenAI, Anthropic, Meta, or Google.
A key part of Nadella’s strategy involves shifting Microsoft from fixed subscription pricing to usage-based, token-driven billing that reflects AI computing consumption. GitHub Copilot has seen strong growth using this model, and products like Copilot Cowork are testing if customers will pay based on actual AI usage.
Microsoft’s Copilot remains central, but workplace adoption has been slower than expected. The company has 30 million Copilot seats compared to over 450 million commercial Office 365 users. Meanwhile, Microsoft’s AI models, led by Mustafa Suleyman’s team, have yet to match the capabilities of OpenAI or Anthropic.
Its relationship with OpenAI has become less exclusive, reducing some early advantages. Nadella is emphasizing a “frontier ecosystem” that works across multiple AI models instead of relying on a single Microsoft model.
The challenge is both economic and technological: Microsoft must demonstrate that AI products can create significant value without causing unpredictable costs for customers. Nadella successfully reinvented Microsoft once through cloud computing and subscriptions; the next test is whether he can do so again with AI.
