Microsoft launches AI powered defence to catch flaws early

AI-powered defence is being framed as a response to an AI-accelerated threat landscape, with Microsoft outlining three core actions to strengthen resilience: integrating advanced AI models into vulnerability discovery and mitigations, enhancing AI-ready posture to reduce exposure, and building AI-powered solutions to defend at scale.

The plan includes embedding models such as Claude Mythos Preview into the Security Development Lifecycle to identify vulnerabilities earlier, and expanding detections via Microsoft Defender and the Microsoft Active Protections Program. It also highlights posture tools like Microsoft Security Exposure Management and Defender External Attack Surface Management to continuously discover internet-facing assets and automate remediation, with Secure Now delivering an integrated guidance-and-action experience.

A multi-model approach is emphasised, with ongoing testing of various providers, and a June 2026 preview for a new AI-driven scanning harness to streamline risk mitigation. These points are presented by Microsoft Security and, according to Microsoft Security Blog.