Google Cloud just rolled out a platform built on a premise that sounds like a sci-fi arms race: using AI to fight AI. The new Google AI Threat Defense, launched on May 27, is an autonomous security system designed to detect, prioritize, and remediate cyber threats at machine speed.
The platform is designed to handle the entire threat lifecycle on its own, from detection through remediation, with continuous monitoring running in the background.
What’s under the hood
Google AI Threat Defense isn’t a single tool. It’s a multi-model AI framework that stitches together several of Google’s most significant security acquisitions and proprietary technologies into one cohesive platform.
At the core sits Gemini, Google’s flagship AI model, handling advanced reasoning and code generation tasks. Wiz, the cloud security company Google acquired, provides contextual risk assessment, helping the system understand which vulnerabilities actually matter versus which ones are theoretical noise. CodeMender handles autonomous code corrections, essentially patching vulnerable code without waiting for a developer to get around to it. And Mandiant, Google’s threat intelligence arm, supplies the incident response expertise and real-world threat data that informs the entire system.
