Microsoft launches Frontier Company with $2.5 billion AI push

Microsoft said Frontier Company will receive a $2.5 billion investment as the software giant tries to move artificial intelligence from pilot projects into day-to-day corporate operations. The new unit will place 6,000 industry and engineering experts inside customer accounts to help companies choose, integrate and scale AI systems that deliver measurable business outcomes and returns on investment.

The company is pitching Frontier as an "outcome-driven engineering organization" built for customers that have already moved past experimentation. Microsoft said the unit will co-design, co-innovate, deploy and continuously improve AI systems at scale, while keeping customers in control of their own data and outputs. Early clients named by the company include LSEG, Land O’Lakes, Unilever and Novo Nordisk, a roster that points to large enterprises with complex workflows rather than small teams testing chatbots.

Judson Althoff, Microsoft’s commercial business chief, framed the launch around a shift in buyer behavior. He said customers now want flexibility to move among models from OpenAI, Anthropic, Microsoft and open-source ecosystems, and said Microsoft made a mistake three years ago by tying Copilot only to OpenAI models. That flexibility is central to the new pitch: Frontier will work with consulting partners including Accenture, Capgemini, EY, KPMG and PwC to help customers mix models and systems around their own business data, not around a single vendor stack. Microsoft says the effort falls under what it calls "Frontier Transformation."

The move also puts Microsoft into a fast-intensifying race to own the implementation layer of enterprise AI. Reuters reported that large corporations are increasingly using a mix of AI technologies instead of depending on one provider, and that Microsoft recently added Anthropic models to Copilot. Amazon Web Services announced a $1 billion embedded-engineer initiative two days earlier, while Anthropic unveiled a $1.5 billion enterprise AI services venture in May with Blackstone, Hellman & Friedman and Goldman Sachs. GeekWire noted that forward-deployed engineering was pioneered by Palantir Technologies, and that the model is becoming a defining tactic across enterprise AI.

The timing matters for investors as much as for customers. Microsoft stock has fallen more than 20% this year, its worst start since 2000, while capital expenditures rose 63% in the most recent quarter to $38 billion and free cash flow weakened. Frontier gives Microsoft a clearer way to argue that its AI spending can convert into revenue from deployment, integration and long-term services, not just model access.