Microsoft is building its own AI models to reduce dependence on outside labs, signaling a strategic shift toward controlling more of its AI capabilities in-house.

What happened: Microsoft unveiled seven new self-built AI models focused on fine-tuning and cost optimization. The company has allocated significant GPU capacity to its internal research and development teams, and during the Build 2026 conference, it showcased plans to help people deploy AI agents and build its own version of OpenClaw.

Why it matters: Until now, Microsoft has relied heavily on frontier labs for AI capabilities. By developing its own models, the company aims to reduce that dependence and gain more direct control over its AI stack layers and applications—a shift that underscores how seriously the tech giant is taking AI as a core strategic asset.

What to watch: On June 4, TD Cowen reiterated a Buy rating on Microsoft and set a $540 price target, citing the company's assertive AI strategy as a reason for confidence in the stock's direction.