Microsoft openly competes with OpenAI and Anthropic, pitches own AI models to enterprises

Microsoft is openly positioning itself as a competitor to OpenAI and Anthropic in the enterprise AI market. During the company's quarterly earnings call on Wednesday, CEO Satya Nadella told Wall Street analysts that Microsoft will sell its own homegrown AI models, agent platforms, and security solutions as lower-cost alternatives, positioning the company as a direct competitor to the premium services that OpenAI and Anthropic are developing for their own growth.

The remarks came as Microsoft posted blockbuster financial results: $90 billion in Q4 revenue with $35.8 billion in net income, and $331.8 billion in full fiscal year revenue with $133.7 billion in net income. Nadella made clear he would not let OpenAI and Anthropic's expansion into applications and agentic infrastructure threaten that cash flow.

Nadella has been warning enterprises to use multiple models rather than relying on any single frontier AI lab. On the call, he escalated that message, arguing that enterprises should maintain a "swappable" model architecture. "The goal is to have the firm be in control of their own destiny," he said. "You got to keep your harness separate from the model ... that means any model at any given time is swappable." He cited last week's Hugging Face breach as evidence that depending on any single model is dangerous.

Microsoft is actively selling its MAI family of models running on its own Maya chips. Nadella announced over a dozen new models across image, voice, transcription, coding, and security, including the company's first reasoning model, MAI Thinking One. For Anthropic's flagship Mythos product, Microsoft earlier this week unveiled MAI Cyber One Flash, which Nadella said "achieves better performance than the much larger Mythos model, but at half the cost when combined with our multi-agent security harness."

Microsoft's strategy is dual: Azure offers over 11,000 models including those from OpenAI, Anthropic, Mistral, and xAI, while simultaneously pushing its own alternatives. The company's core message is that enterprises should use frontier models but not trust them enough to rely on any single one. This "swappable architecture" philosophy is transforming Microsoft from an AI investor into a direct AI competitor, creating growing commercial pressure on both OpenAI and Anthropic.