Microsoft launches $2.5bn AI unit to help firms deploy AI

Microsoft has launched a new business, Microsoft Frontier Company, to help enterprises choose, integrate, and manage artificial intelligence tools from multiple providers.

The company is backing the initiative with a $2.5 billion investment. It will initially work with customers including Unilever, Novo Nordisk, and the London Stock Exchange Group (LSEG).

Instead of relying on a single provider, companies are increasingly combining models from providers such as OpenAI, Anthropic, Microsoft, and open-source developers.

Microsoft Frontier Company aims to simplify that process. It will help businesses select AI models, connect them with internal data, and measure returns on AI investments.

Unlike traditional AI deployments, customers will retain ownership of the systems and outputs created through the service. Microsoft said this approach addresses growing concerns over data sovereignty and intellectual property.

The company will also deploy 6,000 engineering and industry specialists to work directly with customers. These teams will help businesses design, deploy, and improve AI systems based on measurable business outcomes.

Microsoft said the organisation will focus on building two core capabilities. The first is an intelligence platform powered by each customer’s proprietary data and workflows. The second is a trust platform that supports governance, security, compliance, and return-on-investment measurement.

“Three years ago, when we built Copilot, we made a mistake by binding it to OpenAI models only,” Microsoft Commercial Business CEO Judson Althoff told Reuters.

He said enterprise customers now value flexibility more than exclusive access to any single model. According to him, businesses want the freedom to switch between AI providers as technology evolves.

Althoff added that the real value comes from combining AI models with a company’s own data rather than relying on one provider alone.

Although the company remains OpenAI’s biggest commercial partner, it has expanded Copilot to support Anthropic models and other AI technologies.

Industry analysts say enterprises are becoming more cautious about depending on a single AI supplier. Some worry that AI developers could eventually compete with their customers in specialised fields such as software development and legal services.

Patrick Moorhead, CEO of Moor Insights & Strategy, said many businesses want greater control over their AI infrastructure and intellectual property.

Microsoft said Frontier Company will support models from OpenAI, Anthropic, Microsoft AI, open-source projects, and industry-specific providers. Rodrigo Kede Lima will lead the new organisation as president.