Competition in the tech industry is being reshaped around AI agents, and security is increasingly moving with it. Companies including Snowflake are expanding into AI while stepping up security capabilities and governance. Microsoft unveiled an AI agent-based platform for automatically detecting software vulnerabilities and added automated device isolation to its endpoint security product. Workday introduced an independent verification feature for agent deployment, while Cisco strengthened security and governance for local AI agents.

Microsoft used its annual Build developer conference to unveil its own reasoning model, AI agents and cybersecurity tools, signaling a shift toward an AI strategy centered on in-house models. It introduced its first reasoning model, MAI-Thinking-1, and six models for images, voice, transcription and coding. Microsoft AI chief Mustafa Suleyman said the goal is to prove Microsoft can be among the world’s top four research labs.