Microsoft is not announcing a new model. It is not releasing a new API. It is launching a subsidiary whose entire purpose is to help enterprises actually use AI.
Microsoft Frontier Co. will embed 6,000 engineers, consultants, and salespeople directly into customer workflows. The unit starts with $2.5 billion in funding and will work with clients including Unilever and Novo Nordisk. Rodrigo Kede Lima, who previously ran Microsoft‘s Asia business, will lead it.
The key concern behind the move: enterprises are afraid of losing their data and IP to the labs they rent models from. Patrick Moorhead, CEO of Moor Insights & Strategy, put it directly: large businesses suspect that using models from Anthropic and OpenAI will eventually grant these frontier labs expertise to compete with them, especially in fields like coding and law . That’s not a conspiracy theory. That‘s a procurement consideration.
94% of Enterprises Are Extracting No Significant Value
Microsoft’s move is not happening in a vacuum. Amazon committed $1 billion to a similar embedded-engineer initiative two days earlier . Anthropic and OpenAI both established forward-deployed engineering groups in May . Palantir popularized the model years ago.
The convergence suggests the market is moving toward integration, not model dominance. Enterprises are using a mix of technologies — OpenAI, Anthropic, open-source models — tailored to their needs. The trigger is a painful reality: McKinsey found that by the end of 2025, nearly nine in ten enterprises had deployed AI in some part of their business — but 94% were extracting no “significant” value from it . The technology is there. The outcomes are not.
‘We Made a Mistake’ — Microsoft’s Copilot Confession
Microsoft Commercial Business CEO Judson Althoff acknowledged the lesson from the company’s own experience:
“Three years ago, when we built Copilot, we made a mistake by binding it to OpenAI models only. You wanted models to amplify your intelligence and be able to have that sort of swappability for state-of-the-art and fine-tuning.”
That‘s not just humility. That’s a recognition that the market is rejecting vendor lock-in. Enterprises don‘t want to be locked into a single provider. They want the flexibility to switch models as the market evolves. Microsoft’s original Copilot strategy locked them in. Frontier Co. is the correction.
The new unit will help customers select and integrate AI tools — from Microsoft and outside — with their unique internal data. Customers can choose from multiple AI models, including OpenAI, Anthropic, Microsoft AI, open-source, and industry-specific offerings .
P.S. Amazon put $1 billion into the same play two days earlier. Anthropic and OpenAI built similar groups in May. The AI industry is moving from “which model is best” to “who can make models work in the real world.” $2.5 billion and 6,000 employees is Microsoft‘s answer to that question. The answer is not a model. It’s a team. The model race is still happening. The implementation race just got a lot more expensive.
