Microsoft is launching a brand-new operating entity backed by a massive $2.5 billion investment to help corporate clients navigate the increasingly complex and costly world of artificial intelligence.
Dubbed Microsoft Frontier Company, the new unit will embed 6,000 engineering and industry experts directly into client organizations—including early partners like Unilever and Novo Nordisk—to co-design, deploy, and continuously improve tailored AI systems, News.Az reports, citing Reuters.
The move highlights a major shift in how the world’s largest corporations approach AI. Rather than renting out a single AI model from a single provider like OpenAI or Anthropic, businesses are increasingly mixing and matching various technologies, including open-source options, to fit their specific needs. However, this fragmented approach has proven incredibly expensive and slow to yield a real return on investment (ROI).
Microsoft Frontier Company aims to solve this headache by helping companies blend outside AI tools with their own proprietary internal data. Crucially, clients will retain total ownership of the intellectual property generated through this work, rather than feeding that valuable data back into Microsoft’s models.
The strategy also signals a fascinating shift in Microsoft’s own philosophy. Judson Althoff, CEO of Microsoft Commercial Business, admitted that when Microsoft built Copilot three years ago, it made a mistake by binding it strictly to OpenAI models. With competitors like China’s DeepSeek and Google’s Gemini quickly catching up, corporate clients now demand flexibility and the power to swap out different state-of-the-art models as tech evolves.
By launching Frontier Company, Microsoft is stepping directly onto a new enterprise battleground. The tech giant joins the likes of Palantir, cloud rival Amazon Web Services—which recently kicked off its own $1 billion embedded-engineer unit—and standalone deployment ventures from OpenAI and Anthropic, all racing to capture the lucrative market of hands-on corporate AI transformation.