New capabilities to help you secure your AI transformation

Microsoft announces expanded security capabilities to secure and govern AI, including Defender for Cloud’s AI security posture management and runtime threat detections (such as jailbreak attempts, credential theft, and sensitive data leakage), Purview’s AI Hub for AI usage visibility, data protection and non-compliant usage insights, plus new compliance assessments (EU AI Act, NIST AI RMF, ISO/IEC 23894, ISO/IEC 42001). The update also introduces a unified SOC platform experience across Sentinel and Defender XDR, along with broader Microsoft Copilot for Security integrations to provide richer context, automation, and streamlined investigation and response.