Microsoft is openly competing with OpenAI, Anthropic more than ever

Microsoft is in a unique position as AI overtakes the technology industry. It is one of the world’s largest cloud providers and software-as-a-service companies, while also having valuable stakes in the two largest AI labs, OpenAI and Anthropic.

These incentives are starting to conflict as Microsoft posts blockbuster financial results. The company just made an extremely profitable report quarter with revenues of $90 billion and net income of $35.8 billion. For the fiscal year ended June 30, Microsoft reported $331.8 billion in revenue with a net profit of $133.7 billion for the year.

And CEO Satya Nadella has no plans to let that kind of money derail the trajectory of Anthropic and OpenAI — which are expanding into applications and agentic infrastructure that will ultimately allow them to own customer relationships.

Nadella has called on companies to adopt multiple models and stop relying on the advanced AI labs for the agentic harness/app layer.

That’s dangerous, he says, because it forces companies to share too many of their internal secrets with model makers of questionable reliability. He knows his customers. Enterprise IT is afraid of both data leaks and being stuck with a supplier.

Now he’s openly told Wall Street analysts on the company’s quarterly conference call Wednesday that this is an opportunity for Microsoft to sell customers its own homegrown models, along with agents, AI security and more, while promising lower costs.

In other words, he presents Microsoft as an alternative to many of the luxury services that OpenAI and Anthropic are developing for their own growth.

When UBS analyst Karl Keirstead specifically asked Nadella to weigh in on the open versus closed source debate roiling the AI ​​industry, and how Microsoft will benefit from it, Nadella came away with her mouth agape.

“The goal is for the company to take control of its own destiny,” the CEO said of companies. “We are very clear about the architectural design of the platform, which means you have to keep your harness separate from the model… that means any model is interchangeable at any time.”

Of course, Microsoft also sells a menu of suits of armor (also called AI agents) under the name Copilot, including its coding agent GitHub Copilot. Coders are where a large portion of AI dollars are spent today.

And he used last week’s high-profile incident as evidence of his warnings.

“If you even look at the Hugging Face incident, the most important thing to remember from that is that you can’t depend on one model,” Nadella said. “You might need multiple models to solve even some of the problems caused by one model. Like that’s the way to think about it, right? And that means you can’t be subject to the rejection of one model.”

The incident involved an unreleased model of OpenAI breaking out of the sandbox and successfully performing a full hack on Hugging Face, all in a quest to surpass the benchmark. In an attempt to understand what was happening, Hugging Face first tried to use a private border model (which it did not name) which refused to help. So it turned to the Chinese open source model Z.ai GLM 5.2 to analyze logs and defend its infrastructure. The incident has so shocked the industry that even Sam Altman now says AI development may need to slow down a bit.

Nadella also made it clear that Microsoft is happy to sell its own homegrown models, the MAI family, on its own homegrown AI chips, Maya, and offer them as cheaper alternatives.

“Every customer wants the right model for every job based on quality, latency, cost and compliance. We offer the broadest model catalog in the cloud with more than 11,000 models, including the leads from OpenAI, Anthropic, Mistral, xAI, as well as our own MAI family,” he said.

He added, “We’re also accelerating development of our own models. We’ve announced more than a dozen new models across image, voice, transcription, encryption, and security, including our first reasoning model, MAI Thinking One, all with cost-efficient inference at the core for business use cases. We’re co-designing these models with our silicon and we’re seeing a 40% increase in performance per watt when running MAI models on Maya 200.”

As for Mythos? Nadella pointed to Microsoft’s new Mythos competitor announced earlier this week MAI Cyber ​​One Flash. It “delivers better performance than the much larger Mythos model, but at half the cost when combined with our multi-agent security harness,” he said.

Certainly, Microsoft’s CEO says companies should use the frontier models that OpenAI and Anthropic offer in their mix. But his bigger message is: don’t trust them enough to rely on them.