SAN FRANCISCO, July 2 - Microsoft said on Thursday it is establishing a new company to help corporate clients identify and implement artificial intelligence technologies that produce measurable returns for their businesses. The new operating entity, named Microsoft Frontier Company, will begin with $2.5 billion in capital provided by Microsoft to support engagements with customers including Unilever and Novo Nordisk.
The push reflects a shift among large enterprises away from relying exclusively on a single AI supplier. Instead, many firms are assembling a portfolio of models - which can include proprietary offerings, third-party labs and open-source alternatives - and adapting those models to their specific needs. That approach can be expensive and can extend the time before companies realize a return on their AI investments.
Microsoft Frontier Company will work with clients to select and integrate AI tools - whether they originate at Microsoft or elsewhere - and combine them with each customer's internal datasets. A notable element of the arrangement is that customers will retain the outputs of the integration work rather than having those results revert to Microsoft.
The new unit positions Microsoft alongside other firms offering similar services. Palantir Technologies is already deploying Nvidia's open-source models for large customers, while Amazon Web Services has launched its own $1 billion embedded-engineer unit aimed at this market.
Microsoft already maintains partial ownership of ChatGPT-maker OpenAI and earlier this year added Anthropic's models to its Copilot assistant, reflecting strong enterprise demand for those labs' capabilities. Judson Althoff, chief executive of Microsoft's Commercial Business, said the formation of the new firm grew in part from Microsoft's own experience as other models such as China's DeepSeek and Google's Gemini narrowed the performance gap with OpenAI.
Althoff said Microsoft learned practical lessons from the early Copilot deployment. "Three years ago, when we built Copilot, we made a mistake by binding it to OpenAI models only," he said. "You wanted models to amplify your intelligence and be able to have that sort of swappability for state-of-the-art and fine-tuning." He added that, for customers, the combination of data plus models matters more than any single model, and that enterprises need the flexibility to switch among AI models rapidly.
Microsoft Frontier Company will thus aim to reduce integration complexity for major firms and shorten the path to commercial returns by advising on model choice, fine-tuning and data integration while preserving customer ownership of the developed outputs.
The announcement underscores competitive dynamics in enterprise AI services, with major cloud and analytics providers expanding offerings intended to help large customers operationalize a mixture of models and manage costs and timelines associated with tailoring those models to internal systems and data.