It looks like the once celebrated “best bromance in tech” between Microsoft and OpenAI is showing signs of strain. At Build 2025, Microsoft unveiled a slate of new AI announcements that reflect a shifting strategy—one that increasingly involves rivals like Elon Musk’s xAI and European startups, and a more diversified approach to artificial intelligence.
New AI models from xAI and European startups
Among the highlights at Build 2025 was Microsoft’s announcement that it will host a broad portfolio of AI models on its Azure cloud, including new offerings from Elon Musk’s AI startup xAI, as well as European companies Mistral and Black Forest Labs. This expansion increases the total number of AI models available on Azure to over 1,900, underscoring Microsoft’s ambition to provide developers and businesses with a wide range of AI capabilities—all running within Microsoft’s own data centres. This move is designed to ensure reliability and availability at a time when popular AI models are often plagued by outages due to overwhelming demand.
The next-generation AI coding agent
Microsoft also introduced a cutting-edge AI coding assistant during the conference. Unlike previous versions that generated snippets of code based on existing developer input, this new agent can independently take instructions—such as identifying software bugs and outlining strategies to fix them—and carry out the coding and debugging process autonomously. The AI alerts developers once the task is complete for review. This is a significant step forward in AI-powered software development tools and positions Microsoft competitively against OpenAI’s Codex.
Empowering businesses with custom AI agents
Beyond consumer-facing tools, Microsoft rolled out Azure Foundry, a platform that allows businesses to build and deploy their own AI agents tailored for specific tasks within their organisations. These custom agents can be composed from a mix of different AI models, giving enterprises greater flexibility and control over their AI deployments. Azure Foundry reflects Microsoft’s vision for AI to be embedded deeply and flexibly across industry workflows.
Microsoft’s evolving relationship with OpenAI
While Microsoft continues to be the largest investor in OpenAI and provides critical cloud computing infrastructure to power OpenAI’s models like ChatGPT, the partnership has faced challenges. OpenAI’s escalating demand for more computing power and rising operational costs have raised concerns at Microsoft about over-reliance. The fallout from OpenAI’s brief internal leadership shakeup in November 2024 further shook confidence.
In response, Microsoft has been hedging its bets. It invested over $650 million to hire key staff from Inflection, an OpenAI competitor led by former OpenAI co-founder Mustafa Suleyman, and has been integrating rival AI models on Azure. This diversified approach aims to reduce dependency on OpenAI and cultivate a broader AI ecosystem under Microsoft’s umbrella.
What this means for users and the AI landscape
For end users, Microsoft’s moves promise more AI-powered tools integrated across platforms—from smarter software coding aids to bespoke AI agents tailored to business needs. However, this also signals a more fragmented and fiercely competitive AI landscape. The contest for supremacy is no longer just about innovation; it’s about control over infrastructure, access to talent, and strategic partnerships.
The AI race today is a high-stakes battle for power, market dominance, and ultimately, the future of how technology shapes daily life and work. Microsoft’s Build 2025 showcased that the company is playing a complex game—leveraging partnerships and competition alike—to secure its position in this rapidly evolving arena.