Microsoft is working to make more of its artificial intelligence infrastructure self-reliant, even as it continues to collaborate closely with OpenAI, according to remarks tied to a recent discussion involving the company’s AI leadership.

The comments point to a strategy that balances two priorities: preserving a major partnership that has helped power Microsoft’s AI products, while also reducing dependence on any single outside provider over time. Microsoft has invested heavily in OpenAI and integrated its models into products such as Copilot, but the company appears to be broadening its approach as the AI market evolves.

The discussion comes as Microsoft, like other large technology companies, looks to secure more control over the full AI stack. That can include model development, cloud infrastructure, and the tools used to deploy AI systems across consumer and enterprise products. Building more of that capability in-house could give Microsoft greater flexibility in pricing, performance, and product design.

At the same time, Microsoft has continued to emphasize its relationship with OpenAI. The partnership has been central to the company’s AI push, helping it bring generative AI features to a wide range of software and services. The latest remarks suggest Microsoft does not see its work with OpenAI and its own internal ambitions as mutually exclusive. Instead, the company seems to view them as parallel tracks.

That approach reflects a broader pattern in the AI industry. Major cloud and software companies are trying to avoid overreliance on outside model providers by investing in their own systems, while still using external partnerships to accelerate product development. For Microsoft, that could mean maintaining access to leading models from OpenAI while building the ability to support more of its AI offerings independently.

The shift also has strategic implications for Microsoft’s cloud business. As AI demand grows, control over infrastructure and model access can influence how effectively a company serves enterprise customers. A more self-sufficient stack could help Microsoft manage costs and performance at scale, while also giving it more leverage as it competes with other AI and cloud providers.

The comments do not suggest that Microsoft is preparing to end its OpenAI relationship. Rather, they indicate a company trying to hedge its bets in a fast-changing sector. OpenAI remains an important partner, but Microsoft appears to be preparing for a future in which it can rely on a broader set of internal capabilities.

For now, the message from Microsoft is one of expansion rather than replacement. The company is still working with OpenAI, but it is also signaling that it wants more of the AI stack under its own control as the technology becomes more central to its products and business strategy.