Microsoft brings more AI to PCs

Microsoft is accelerating its efforts to bring artificial intelligence directly to personal computers, introducing a new AI coding model and security technology designed to support AI agents on Windows devices. The developments reflect the company's broader strategy to reduce dependence on cloud computing and make personal computers capable of handling increasingly complex AI tasks locally.

At an event in San Francisco, Microsoft introduced Microsoft Execution Containers, a technology designed to restrict AI agents from accessing files, data, and system resources without authorization. According to Microsoft, organizations will be able to establish rules governing AI agents, while Windows will enforce those restrictions on individual devices. Anthropic, OpenAI, and Nvidia are expected to use the technology. Microsoft CEO Satya Nadella emphasized the importance of security as AI agents become more capable of independently performing tasks. The company argues that personal computers must provide a controlled environment where AI systems can operate without gaining unrestricted access to sensitive information. Microsoft also highlighted its growing partnership with Nvidia. The two companies have worked to enable demanding AI applications, including software development and complex business tasks, to run directly on desktops and laptops using Nvidia's AI-focused chips. Microsoft introduced the Surface Laptop Ultra, featuring Nvidia technology, as part of this strategy.